+ All Categories
Home > Documents > Contributions to Drilling, Completion and Workover ...Horizontal drilling Casing drilling ERD. 7 2....

Contributions to Drilling, Completion and Workover ...Horizontal drilling Casing drilling ERD. 7 2....

Date post: 09-Jul-2020
Category:
Upload: others
View: 25 times
Download: 0 times
Share this document with a friend
148
1 Contributions to Drilling, Completion and Workover Technology Beiträge zur Bohr-, Komplettierungs- und Aufarbeitungstechnik HABILITATIONSSCHRIFT zur Erlangung der Lehrbefugnis für das Fachgebiet Drilling, Completion and Workover vorgelegt von Dr. Dr.-Ing. Catalin Teodoriu aus PLOIESTI, Rumänien genehmigt von der Fakultät für Energie- und Wirtschaftswissenschaften der Technischen Universität Clausthal 11.06.2011
Transcript

1

Contributions to Drilling, Completion

and Workover Technology

Beiträge zur Bohr-, Komplettierungs- und Aufarbeitungstechnik

HABILITATIONSSCHRIFT

zur Erlangung der Lehrbefugnis

für das Fachgebiet Drilling, Completion and Workover

vorgelegt von Dr. Dr.-Ing. Catalin Teodoriu aus PLOIESTI, Rumänien

genehmigt von der

Fakultät für Energie- und Wirtschaftswissenschaften

der Technischen Universität Clausthal

11.06.2011

2

Für Lavinia

3

Content

Preamble 3

1. Introduction 4

2. Selected contributions to drilling and production 6

2.1. Advanced modeling of the stress/strain state in tubular goods 6

2.2. New developments in deep drilling technology 9

2.3. Experimental investigation of multiphase flow 10

2.4. The importance of recovery factor 12

3. Comments on selected categories of peer-reviewed papers and comparable contributions 14

References 22

Appendix 1: List of papers and other publications

Appendix 2: List of papers grouped by subject

Appendix 3: List of selected papers for this habilitation (full-length papers are presented)

4

Preamble

Publishing is the key in characterizing and defining the academic development of a faculty member. The author’s peer-reviewed and conference papers are listed in Appendix 1. They can be grouped into two primary categories: papers dealing with investigations in the areas of drilling and completions, and papers with a focus on production and workover operations. Each category is then subdivided into the following topics:

Drilling and completions

Equipment, processes, special procedures (i.e. horizontal drilling) Critical equipment and components Analysis and modeling of operational procedures, well control methods Drilling and cementing fluids Health, Safety and Environment (HSE) aspects

(thread compounds, fluid rheology, formation damage) Advanced drilling technology Geothermal technology Advanced teaching methods and tele-teaching

Production operations and workover • Equipment to improve recovery efficiency, multiphase flow in pipelines • Tubular stress and strain, tubular integrity over field life,

sealability of tubular components, temperature effects, corrosion issues • Liquid loading in gas wells • Recovery of hydrocarbons • Optimization of geothermal energy recovery

The full list of papers, organized according to the scheme presented above, is presented in Appendix 2. From the total number of publications listed in Appendix 1 (94 in total, plus 2 pending peer-revision), only those publications listed in Appendix 3 were chosen for the application for habilitation at Clausthal University of Technology.

5

1. Introduction Drilling and production are at the heart of the petroleum engineering discipline. Together, they cover fundamental hardware and operational aspects of the upstream sector of the Exploration and Production (E&P) business. Several advanced technologies have been developed since the onset of hydrocarbon exploitation in 1859, both in the USA (Titusville, Pennsylvania) and Germany (Wietze, Nieder-sachsen).

Drilling technology and hydrocarbon production operations have remained key areas in the education and training of petroleum engineers for more than 100 years.

With a specialization in mechanical and process engineering within petroleum engineering, the author is convinced that there is still a strong need to further improve components and systems in drilling and production operations. Figure 1 gives an example of the continued development of drilling technology (specifically with respect to mechanization of pipe handling at the rig floor) throughout the years, starting from the 1970s.

All drilling and production operations require a thorough understanding of the associated environmental aspects. Traditionally, the majority of accidents take place close to where drilling operations are carried out and in particular where heavy hardware components are to be maneuvered during tripping or make-up of downhole assemblies.

A few years back, Norway initiated an investigation of situations where personnel safety was endangered. It was concluded that human work should be replaced by machines wherever the majority of accidents were proven to take place, e.g. in the vicinity of the borehole. This theory soon spread worldwide, first to deep-sea drilling rigs, and later to other drilling installations. Experience with automated systems quickly showed that not only safety is increased, but also the overall drilling time is reduced compared to manual operations. Additionally, automated systems to make-up Rotary shouldered connections are much more accurate than the traditional make-up with chains and rig tongs, thus reducing connections damage. State-of-the-art automated make-up devices also provide sufficient data to analyze the quality of the connection make-up, which can be used to increase the connections lifetime and reduce drill string failures.

Drilling as well as production operations are an integral part of hydrocarbon exploitation as a whole, and therefore can not be studied in isolation. This is the reason behind the broad range of topics covered by the selected papers and reports.

6

Figu

re 1

: E

xam

ples

of c

ontin

ued

adva

nces

in d

rillin

g te

chno

logy

, with

a fo

cus

on m

echa

niza

tion

of p

ipe

hand

ling

at th

e rig

floo

r.

Spin

ning

chai

nre

plac

edby

pipe

spin

erTe

leco

intr

oduc

edsi

mpl

e M

WD M

ultis

enso

r M

WD

1 in

trod

uced

Sper

ry‐S

un

intr

oduc

edLW

D 2

MH

z

Part

ially

mec

hani

zed

pipe

hand

lingBa

ker H

ughe

s in

trod

uced

inte

grat

edse

rvic

esbi

t+m

otor

+MW

DSc

hlum

berg

erin

trod

uced

IDEA

L sy

stem

with

inst

rum

ente

dm

otor

Firt

soni

cto

ol

Com

plet

ely

mec

hani

zed

pipe

‐han

dlin

gD

ual/

Tri‐A

ctD

erri

ck

Hig

h ra

te m

udte

lem

etry

Off

shor

e H

igh

Pres

sure

Hig

h Te

mpe

ratu

reD

rilli

ng (H

PHT)

Topd

rive

East

man

& S

mith

in

trod

uced

stee

rabl

em

otor

s

Pow

erPa

km

otor

s

Step

chan

gein

Q

HSE

Schl

umbe

rger

intr

oduc

edAR

DC5

toolRo

tary

Stee

rabl

eSy

stem

sH

oriz

onta

l dr

illin

g

Casi

ngdr

illin

gER

D

7

2. Selected contributions to drilling, completion and workover

The following topics have been selected for this habilitation synopsis. These topics cover drilling, completion and workover aspects of petroleum engineering with a special focus on equipment.

• Evaluation of wire line drill pipe with double shoulder connections and suggestions for further improvements.

• Study of the influence of contact pressure within the threaded turns and shoulder part of tool joints

• Determination of the lifetime of casing couplings and development of an analytical solution

• Testing of thread lubricants to enhance the seal capacity of API connections • Study of excessive loads on large size anchor pipe underneath the casing well

head • Discussions of new developments in deep drilling • Multiphase flow with sand transport in horizontal pipes • Development of advanced testing facilities for multiphase flow investigations • Close view on the low recovery factor in many oil fields

and discussions on possible improvements

In what follows, an insight into the author’s work on stress/strain state in tubular goods, deep drilling technology and multiphase flow is presented.

2.1. Advanced modeling of the stress/strain state in tubular goods

The advances in petroleum technology are linked to advances in the fundamental sciences, such as mathematics, physics and material engineering. For example, the need to develop more resistant downhole tools (stronger threaded connections in particular) led to the need for the petroleum discipline to implement mathematical techniques (e.g. numerical solvers) and materials characterization techniques (e.g. the Finite Element Method (FEM)). A brief overview of these developments is presented below.

In the analysis of shouldered connections, analytical methods offer, in comparison to other types of investigation (DMS, FEM, photo-elasticity), the advantage of reduced computing time and great flexibility in varying influencing parameters. Thus, these methods allow easy change of input parameters and/or geometry (configuration of shoulder area, thread shape and cone angle) in the calculations workflow. A time scale of key developments in the area of stress distribution in threaded connections, including analytical models and make-up torque criteria, is presented in figure 2.

8

Figure 2: Milestones in the developments of analytical models for stress distribution in threaded

connections

Finite element analysis and other experimental research are applied specifically to a given type of threaded connection, and therefore not of general applicability, but they can accurately define local stress distribution and help to improve thread features.

All the analytical models available to compute stress distribution in threaded connections assume that the thread flank stress distribution is uniform. Thus, the resulting forces are considered as acting on the mid-axis of the thread tooth. Usually, the mid-axis of the thread tooth is considered to be identical to the pitch radius. As presented in several papers (Marx et al., 2001; Sager, 1986; Teodoriu, 2003; Birger, 1944; Ulmanu, 1973) the thread tooth deflection strongly influences the stress distribution computation. Figure 3 shows a qualitative comparison between a connection with elastic thread teeth and one with rigid teeth. The thread tooth rigidity changes the stress distribution in the connection and the tooth loading, which may have a negative influence on the connection reliability. To compute the connection deformation induced by the tooth deflection, Paland used in its paper (Sager, 1986) a beam model with distributed load (see figure 4). A few years later, Marx et al. (2001), Sager (1986) and Teodoriu (2005) initiated the first discussion group related to the force position on loaded flank. It was stated that the force position may be the key to understand the stress distribution in threaded connections.

1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Maduschka (1937)

Ulmanu (1973)

Sager (1986)

Kwon (1990)

Marx et al. (2001)

Teodoriu (1999)

Birger (1944) Paland (1967)

Brinegar (1970)

Farr (1957)

Van der Wissel (1959)Teodoriu (2003)

9

The casing strings are made of pipes of a certain length (e.g. 10 m) which are connected by means of a threaded coupling. Actual standards only account for static loads for the casing string. Special cases are represented by steam injection and geothermal wells, where variable loads are applied to the casing string due to temperature and internal pressure variations. Due to movement restriction (i.e. cement ring around the casing) temperature variations will induce thermal stresses inside the casing string. The value of induced thermal stresses may overcome the material’s yield strength, and the fatigue behavior of the casing material can be classified as low cycle fatigue. The presence of geometry changes in the casing body (i.e. threads) will amplify the local stress distribution, and will reduce the low cycle

Figure 3: Stress distribution for a connection with rigid teeth (dotted line) and with elastic teeth (continuous line)

Distance from shoulder

Axia

l stre

ss

Figure 4: Schematic simplification of a loaded flank

Flank load

Active flank

Un-loaded flank h

10

fatigue resistance. The local stress concept permits the evaluation of low cycle fatigue resistance of such casing connection using the experimental results on uniaxial small-scale specimens and reduces the time and costs needed for classic statistical evaluation on full-scale specimens. The implementation of this concept makes use of analytical and numerical methods. FEMs are intensively used to quantify the local stress distribution and the local concentration factors. Once these parameters are determined, they will be used in a semi-analytical method to determine the fatigue resistance of the casing and its couplings.

The same analytical and numerical methods can also be used to provide new insights into connection resistance. For instance, the numerical methods can be used to validate the analytical solutions, hence reducing the need for full-scale testing.

2.2. New developments in deep drilling technology

Rotary drilling became the preferred deep drilling method for oil and gas wells during the mid- and late 20th century. It is described as the method where the rock is crushed by simultaneously developing a continuous circular motion of the bit (rotational speed), pumping a drilling fluid down the drill pipe to flush out the debris (flow rate), and exerting a downward force on the drill bit to enforce the rock breakage (Weight On Bit (WOB)). The drilling system used to achieve this can be either a Kelly/Rotary Table or a Top Drive system. Prior to the 1980’s, drilling with the Kelly was a common trend, both onshore and offshore, therefore the actual technology was called: Rotary drilling = drilling using a Rotary table system. The Kelly was basically used to transmit the rotational motion given by the Rotary table to the drill pipes, and the fluid circulation was performed by means of mud pumps connected to the Kelly through a swivel that allowed free rotation of the Kelly itself. Moreover, drilling using a Kelly allowed only drilling with singles (i.e. only one drill pipe length at a time was made). After the 1980s, the introduction of the Top Drive system revolutionized the way of drilling wells. This system enabled the simultaneous tripping of the drill string and circulation of the drilling fluid. In this case, the making up of doubles, triples and quadruples was made possible, which radically increased drilling speed (but not necessarily the rate of penetration) and allowed for deeper wells. In both systems, downhole motors could be used and WOB was applied statically by the drill collar (representing two thirds of the DC length) and dynamically by raising and lowering the Drilling Assembly. Top Drive systems became the preferred technology for offshore drilling activities.

The impact of mechanized rig operations on work efficiency is proven by many papers (Teodoriu, 2005). A wide spectrum of power tongs can allow mechanized operations at the rig floor. Mechanized rig operations improve the reliability of oilfield tools. When Rotary drilling is used, the most important components to characterize this process are the tubular goods and their connections, which can be achieved by using power tongs. The main purpose of a power tong is to perform an optimum

11

make-up of the threaded connection between two tubulars. Accurate make-up of Rotary Shouldered Connections (RSC) is of greatest importance towards achieving the best torque and axial force transmission through the connection, which optimizes the connection’s lifetime under downhole conditions. The make-up torque value of RSCs depends on the friction coefficient between the threads and the parts of the shoulder that are connected during the make-up procedure. This friction coefficient normally cannot be individually measured or determined; this is why the API Recommended Practice 7G gives an average recommended make-up torque based upon an assumed friction coefficient, adding a certain safety margin. Thus, the majority of RSCs are being made-up with a torque value that differs from the connection optimum torque. This has two consequences: the lifetime of each connection is not maximized, and the torque and axial force values that can be transferred through the connection are different from the technical maximum, which the connection would be able to transfer if it was made-up with the optimum make-up torque. The make-up of drill string components (such as Drill Pipe, Heavy Weight, Drill Collars and Bottom Hole Assemblies) is a key task in the drilling process, which requires significant non-productive time.

2.3. Experimental investigation of multiphase flow

The reservoir fluids produced from oil wells are rarely pure liquid or gaseous hydrocarbon mixtures. Most often, the fluid emerges as a multiphase mixture of natural gas and oil, but, in many systems, water is present as well as a variety of solid phases (sand, hydrates, and asphaltenes). Traditionally, the flow rates of well fluids have been measured by separating the phases and measuring the outputs of the separated fluids by conventional single-phase techniques (Falcone, 2009). However, the economics of offshore oil recovery have moved towards subsea completions with multiphase pipelines over long distances to either the shore or to existing platforms. The problems of multiphase production through wellbores and pipelines require therefore an in depth understanding through fundamental research.

As tubular guarantee the flow from the reservoir to surface, they must be designed based not only on loads (axial load, internal pressure, etc.), but also on the flow regime that is expected through them. Multiphase flow modeling requires extensive experimental investigations in order to validate analytical and numerical models. Such investigations can be performed in dedicated experimental testing facilities capable of mimicking multiphase flow phenomena.

At the onset of the petroleum industry, oil and gas reservoirs were discovered on land. However, at the end of the 19th century, reserves discovered near shore initiated the offshore petroleum industry. In more recent times, economics and demand have justified exploration and development of fields offshore in the Gulf of Mexico, the North Sea, West Africa, the northeast coast of Brazil, Australia and other regions around the world. For each field, there are several development options to bring production to the market. The successful design of an offshore development option depends on many factors, such as water depth, surface/underwater conditions

12

and available infrastructure, which lead to the choice of a fixed platform, a floating vessel, a floating structure or a subsea system. For the final hydrocarbons to be placed on the market, phase separation is necessary to remove unwanted water or solids and provide the right specifications for the oil (e.g. API gravity) and for the gas (e.g. calorific value). A production system must be configured so as to guarantee flow assurance from the reservoir to the point of sale.

Sand management and multiphase production technology has become conventional for the exploitation of a vast amount of petroleum resources trapped in unconsolidated formations located in offshore, deep water and ultra deep water environments. Increased water depths create a requirement for reliable subsea wells and flowlines to minimize flow assurance issues. Modeling of three-phase gas-oil-sand flow in horizontal wells remains a challenge to the petroleum industry. One effect of the increased frictional pressure losses in horizontal wells is the reservoir performance. The effect of increasing pressure loss in horizontal wellbore translates to an increase in reservoir pressure drawdown.

The advantages of sand management have been identified by a large number of investigators (Salama, 1998; Dusseault et al., 1998; Dusseault and El-Sayed, 2001; Dusseault et al., 2000; Geilikman and Dusseault, 1997; Bratli et al., 2000; Tronvoll et al., 2001; Dusseault et al., 2002; Bello and Fasesan, 2003; Bello and Fasesan, 2004). A major challenge in sand management during production is the understanding of solid particle hydrodynamic behavior, which is related to the profile of sand velocity, the interaction between the suspended solid phase and other phases, and the gas-liquid turbulent slug flow structure. The understanding of the behavior of the suspended sand particles during gas-liquid multiphase production and transfer operations enables the assessment of the severity or risk of potential sand deposition. It also helps to identify the parameters that would need to be controlled in order to check sand deposition and optimize production.

Inefficient sand transport by the gas-liquid flow can lead to numerous problems, including sand deposition and accumulation in the well and in the production lines, increased pressure loss, increased erosion risk, frequent and expensive cleaning operations, and increased downtime (Oudeman, 1993; Appah and Ichara, 1994; Appah et al., 1997).

Although many studies have been performed on the transport velocity of various solid particles during production of manganese nodules (Sakaguchi et al., 1987a; Sakaguchi et al., 1987b; Sahara and Saibe,1991; Sakaguchi et al., 1992; Sakaguchi et al., 2000). However, the behavior of solid particles during production of manganese nodules is essentially different from oil-gas-sand multiphase production and pipeline transportation. The oil-gas-sand multiphase pipeline transportation consists of fine particles while in manganese nodules production coarse particles are contained. The size of the fine particles is usually less than 0.7 mm and that of course particle is about 5 mm. These particularities require a rethinking of testing

13

facilities and measurement concepts, which have been implemented in the ITE Multiphase Flow Loop system.

To efficiently design and operate gas-liquid-sand flows in wells and pipelines, the degree of the distribution of the solid phase in the tubular must be quantified and controlled. Because it is crucial to understand the influence of various operating and system conditions on the solid velocity profile, a measuring technique capable of probing the internal flow structure without being invasive to it is highly desirable. The implementation of digital image techniques is an established method for multiphase flow visualization and analysis (Caicedo et al., 1993; Sam et al., 1995; Kundakovic and Vunjak-Novaakovie, 1995; Kaftori et al., 1995a; Kaftori et al., 1995a; Nino and Garcia, 1996; Aloui and Souhar, 1996; Gopal and Jepson, 1997; Marques et al., 2002; Leifer et al., 2003; Shen et al., 2004). It has advantage over other existing solutions for solid velocity and hold-up measurements because of its non-invasive character. The fast response of the technique justifies its likely application to multiphase process control. Hence, the development of a non-intrusive measuring technique that can provide reliable information on the solid phase flow structure appears to be key towards the development of a mechanistic and hydrodynamic approach to sand transport in multiphase production. This may improve operations and safety, and reduce capital costs.

To respond these new challenges new flow loops have been re-designed and built. A thorough investigation and ranking of world-wide flow loops for multiphase flow experiments should include all of the following factors: loop geometry, dimensions, operating pressure and temperature, range of phase flow rate, equipment and instrumentation, piping material, fluid properties, data acquisition and information processing systems. However, the objective of the research task is to illustrate how to approach such an investigation and to identify future needs for niche experimental investigations. The outcomes of the performed review show that there is a need for a dedicated flow loop to mimic the dynamic interactions between reservoir and wellbore. Furthermore, a dedicated flow loop to understand sand transport phenomena in multiphase flows is also required.

2.4. The importance of the recovery factor

Why investing so much time in research? The oil and gas business is constructed around the amount of fluids stored in the reservoir. The actual recovery factors are rather low for oil (average of 35%) and they can be increased through new technology and extraction concepts. Research activities performed within academia allow, through technology transfer, a better exploitation of our finite resources, and will also lead to smooth transition to renewable and environment friendly energy resources.

High-pressure and high-temperature (HPHT) gas reservoirs are defined as having pressures greater than 10,000 psia and temperatures over 300ºF. Modeling the performance of these unconventional reservoirs requires the understanding of gas

14

behavior at elevated pressure and temperature. An important fluid property is gas viscosity, as it is used to model the gas mobility in the reservoir that can have a significant impact on reserves estimation during field development planning. Accurate measurements of gas viscosity at HPHT conditions are both extremely difficult and expensive. Thus, this fluid property is typically estimated from published correlations that are based on laboratory data. Unfortunately, the correlations available today do not have a sufficiently broad range of applicability in terms of pressure and temperature, and so their accuracy may be doubtful for the prediction of gas viscosity at HPHT conditions. Ehsan et al. (2009) performed a sensitivity analysis to assess the effect of gas viscosity estimation errors on the overall gas recovery from a synthetic HPHT reservoir, using numerical reservoir simulations. The result shows that a -10% error in gas viscosity can produce an 8.22% error in estimated cumulative gas production, and a +10% error in gas viscosity can lead to a 5.5% error in cumulative production. These results indicate that the accuracy of gas viscosity estimation can have a significant impact on reserves evaluation.

Experimental investigations on gas viscosities at HPHT conditions is not an easy task and require development of laboratory equipment and testing procedure that ensure first of all the safety of the experimental work. Ehsan et al. (2009) show an example of the work performed on gas viscosity measurements where a falling body viscometer was used to measure the HPHT gas viscosity in the laboratory. The instrument was calibrated with nitrogen and then, to represent reservoir gas behavior more faithfully, pure methane was used. The subsequent measured data, recorded over a wide range of pressure and temperature, were then used to evaluate the reliability of the most commonly used correlations in the petroleum industry. The results of the comparison suggest that at pressures higher than 8,000 psia, the laboratory measurements drift from the National Institute of Standards and Technology (NIST) values by up to 7.48%.

15

3. Comments on selected categories of peer-reviewed papers and comparable contributions

In what follows, a summary of the selected papers based on the categories shown in chapter 2 is presented.

3.1. Evaluations and proposals for further product improvements on wire line drill pipe with double shoulder connections

The wire line technology uses drill pipes having threaded connections with no external or internal upsets. These connections are usually weaker than the pipe body and require high make-up torque values. To achieve this, double shoulder connections are used because they have the advantage of improving the stress distribution on thread turns.

The paper in question shows the analytical, numerical and experimental work performed on a double shoulder connection in order to understand the effect of the internal shoulder as well as the shoulder load distribution. To solve this, the existing equation for shoulder connection calculation was extended to incorporate the internal shoulder effect. The results showed a good comparison of the analytical, experimen-tal and numerical results. The extended equation to calculate the make-up torque can be easily generalized for any double shoulder connection.

The technology of double shoulder connection has a high impact on the future of drilling technology, especially when deep drilling activities like HPHT and geothermal are involved.

Untersuchung zur Belastbarkeit von SLIMHOLE-Bohrgestängen mit Doppelschulter, Erdöl, Erdgas, Kohle, 125 Jg. 2009, Heft 7/8.

3.2. Study of the influence of contact pressure within the threaded turns and shoulder part of tool joint

The previous work showed the importance of combining numerical and experimental work in order to better describe the load envelope of a given connection, in particular double shoulder connections. The research work is related to the second selected paper is enhancing the fundamental studies on threaded connections by analyzing the influence of contact pressure within the thread turns and shoulder part of a tool joint.

Drill pipe connections are a primary component of the drill string, and the entire “well security” depends upon tool joint performance reliability. An adequate connection between two drill pipes depends on the quality of the assembly process, which is significantly affected by thread compound performance. Since the variety of thread

16

compounds is large, standards have been developed to determine thread compound performance and to define minimum thread compound properties. Thread compound frictional performance is normally defined by its friction coefficient (COF) in the standard American Petroleum Institute (API) Rotary shouldered connection (RSC), or the friction factor (FF) relative to a standard API friction coefficient, which is accepted to be 0.08. The 0.08 value was determined by the API after multiple tests utilizing standard API RSCs and metal-based thread compounds, primarily 40% zinc and 60% lead. The friction coefficient of a thread compound has been generally considered to be constant. Although the API defines a constant coefficient for a given load range, recent studies indicate that the friction coefficient can vary considerably depending on connection geometry, contact stress and thread compound composition.

As stated by Farr (1956), the make-up torque of a connection is a function of its geometry, the pre-defined buck-up force and the friction coefficient between the two members of the connection (pin and box):

⋅+⋅+

⋅⋅= ssth

pv R

RPFM µµβπ cos2

. (1)

Typically, under field conditions, the applied make-up torque M can be measured, while the buck-up force Fv can only be estimated. As equation (1) shows, knowing the applied torque M and the parameters within the brackets, the Farr formula can be used for reverse calculation of the induced buck-up force in the connection. Since a better understanding of the friction process and consequently an accurate make-up torque calculation allow optimizing the load resistance of the connection, more research must be focused on friction coefficient behavior under real conditions (e.g. temperature, contact pressure, composition, etc). Based on the findings presented in this paper as well as in previous work, a DGMK Project was initiated.

Rotary-Shouldered Connections Make-up Torque Calculation Considering the Effect of Contact Pressure on Thread Compound’s Friction Coefficient; OIL GAS European Magazine 4/2009.

3.3. Determination of lifetime for casing couplings and development of a method for an analytical solution

The success of any drilling operation is to drill the well to the target depth (TD) in a safe manner. This includes the integrity of the drill string (i.e. drill pipes and their threaded connections) as well as the integrity of the wellbore, especially of the casing string. Well lifetime definition has changed during the history from few years to as long as 50 years. This generates new research areas, one of them being the topic presented in the following.

17

Any well drilled deeper than 3 km would present temperatures in line with those typical of higher enthalpy geothermal resources. However, this paper focuses on well integrity issues for hot wells, independent of their depth. In Europe, the majority of the geothermal wells, which have been drilled to depths greater than 4000m, are completed with surface casing diameters of 18−5/8in. (473mm) or larger. The large diameter of production casings is a consequence of the amount of fluids that needs to be produced from geothermal wells. For large installed power systems, production diameters of 13−3/8 in. (340mm) are required, but such diameter requirements strongly affect well costs. Over the operating life of a well, its casing string is generally subject to external loads that can be considered as static or quasi-static. Current industry design standards, such as those issued by the American Petroleum Institute (API), consider the casing string to be statically loaded, yet it can be subject to variable loads due to changes in temperature or internal pressure during geothermal operations. As the movement of the casing is restricted by the presence of a cement sheath, temperature variations induce thermal stresses in the casing string, which may exceed the yield strength of the casing material. Thus, the fatigue behavior of the casing material during the operational life of a well can be classified as low-cycle fatigue (LCF). The presence of geometrical variations in the casing body such as the connection threads will amplify the local stresses and reduce the LCF resistance of the casing. The surface casings of deep geothermal wells are exposed to significant temperature variations during drilling, which may affect their subsequent integrity. The size of the surface casing depends on that of the production casing and also on the drilling challenges presented by that particular well. The theoretical and experimental work focuses on the fatigue resistance of an 18−5/8in. (473mm) diameter casing with Buttress thread connections, which is a common size for surface casing in geothermal wells, as reported by several authors.

Extensive Finite Element runs have been used to determine the local stress values due to stress concentration effects. Figure 5 shows the stress distribution for two different load types (tension and compression). It shows a slightly different behavior of the connection between the two load scenarios.

Based on the FEM analysis, it was found that for a buttress connection the stress concentration factor, K, differs from tension to compression. This variation can be explained by the more aggressive bending on the thread turn for the tensile load and also by the asymmetric geometry of the Buttress thread that has different flank angles. The following results were obtained: K = 3.51 for tension, K = 2.73 for compression. In reality, the incomplete thread turns present sharp edges and therefore higher stress concentration factors may be expected.

Comparing completion design in hydrocarbon and geothermal wells: the need to evaluate the integrity of casing connections subject to thermal stresses, submitted to the Geothermics journal in April 2008.

18

-702.391

-581.721-461.052

-340.383-219.714

-99.04521.624

142.294262.963

383.632

SY (AVG)RSYS=0DMX =.169922SMN =-702.391SMX =383.632

MN

-545.063

-413.808-282.552

-151.297-20.042

111.213242.469

373.724504.979

636.234

TIME 1000SY (AVG)RSYS=0DMX =.153252SMN =-545.063SMX =636.234

Figure 5: Stress distribution at the thread root for the compression (top) and tension

(bottom) load cases, as simulated with FEM. High stresses at the thread root are shown in red.

3.4. Study of excessive loads on large size anchor pipe underneath

the casing well head

Since modern wells are drilled to reach complex purposes, loads considered for casing design are in some cases extreme. HPHT wells are known for their extreme environment and therefore the acting loads on the casing string. The casing design under axial compression was learn from steam injection projects. For these cases there was found that pre-stressing the casing string may reduce the negative effect of axial compression. During more than 40 years of steam injection it has been found that Buttress connections have a good loading capacity under high axial tension loads. The advantage of a Buttress connection is given by its simplicity and low costs comparing to premium connections. Therefore, it is usual for some casing strings like anchor casing to have Buttress connections, which are cheaper, especially for large casing diameters. Concerning anchor casing the acting loads are typically low, but at the surface the entire casing strings are hanging on it. It is necessary to investigate the compression resistance of such casings. The threaded connections are mostly designed to have a maximum tension resistance. There are some new connections on the market especially designed for high compressive loads, but for special purposes like drilling with casing or high temperature wells. This special condition makes the price for such connections high. The paper presents a method to quantify the compression resistance of Buttress connections based on experimental and

19

analytical results. Experimental, numerical and analytical methods have been combined.

Figure 6 shows a comparison of the full scale test versus Finite Element results. Both methods pointed out that the system will fail due to a local buckling of the last engaged threads.

Buttress Connection Resistance under Extreme Axial Compression Loads, Oil and Gas Magazine, 4/2005, Volume 31, ISSN 0342-5622.

Figure 6: Measured and estimated failure modes for a Buttress connection.

3.5. Testing thread lubricants to enhance seal capacity of API connections

This paper presents a combination of “near praxis” and laboratory development. Petroleum Institute are generally criticized for their lack of “fundamental research” due to their praxis-oriented approach. This work presents the experimental results of tests carried out on four different compounds using the improved “grooved-plate” method. The tests have shown a large variation of the tested-thread-compounds sealing capacity. Starting from the experimental results and the theoretical analysis of the American Petroleum Institute (API) connection, a useful chart was built to determine the real connection resistance, based on its initial make-up torque, see figure 7. The chart offers to engineers involved in the design of a fracturing process the possibility to estimate the maximum pressure that may lead to a connection leak. The paper combines experimental and numerical approaches to solve an industry-known problem: thread compound characterization.

The tests performed on four different types of compounds have shown that the API compound has the lowest leak resistance in conjunction with the API-thread type. The Buttress leak resistance has an asymptotic behavior. At contact pressures higher than 100 MPa, the leak resistance is constant.

The difference between API-round and Buttress leak resistance consists in the contact-pressure dependency of the API-round leak resistance. It is recommended for API-round connections to be made-up with optimum make-up torque or higher.

Possible buckling zone

Measured buckling

20

As a result of this study, we recommend that research focused on the relationship between dope-rheological properties, time, and flow behavior should be performed.

Sealing Capacity of API Connections - Theoretical and Experimental Results, SPE 106849, March 2009 SPE Drilling and Completion Journal.

Figure 7: Chart for API connection leak resistance estimation knowing the thread compound leak resistance.

3.6. Discussion of new developments in deep drilling

It has already been presented above that focal research points are moving throughout the history of drilling. This is why reviews must be performed in order to identify new tendencies in the drilling business.

Such developments are to a large extend related to the complexity of the reservoirs that have to be drilled. New drilling rigs are introduced onto the market to enhance safety and working conditions, while downhole technology becomes more complex. Nevertheless, new developments include some improvements, especially in the area of threaded connections. Two papers shall be mentioned here with respect to the discussion upon new developments: Neue Entwicklungen in der Bohrtechnik, Akademie der Geowissenschaften zu Hannover Veröffentlichungen, 2005, Heft 25, ISBN 3-510-95943-4 and Increasing Geothermal Energy Demand: the Need for Urbanization of the Drilling Industry, Bulletin of Science, Technology & Society 2008, Volume 28, No. 3, Special

21

Issue, Part 2: The Age of Alternative Energy: Is the Future Renewable?, DOI:10.1177/0270467608315531, 16 April 2008.

Current drilling practices are based on the principle of Rotary drilling, where a dedicated rotating system drives a cutting tool (i.e. drill bit) in order to break the rock. A continuous weight is applied to the bit to provide the necessary cutting force. The weight on the bit is controlled via the hoisting system, which is also used to lift and trip the well tubular in and out of the well. The cuttings are removed from downhole and transported to surface by a drilling fluid (air, foam or drilling mud). Cuttings removal is the main function of the circulating system that allows the drilling fluid to be pumped downhole (via the drill pipe) and returned to surface (via the annular space between drill pipe and hole), where it is reconditioned. Until recently, the oil and gas drilling industry portrayed an image using equipment and working environment that was “large, dirty and downright ugly”. Nowadays this is unacceptable and the industry pays close attention to making drilling rigs clean, safe and environmentally friendly. Consequently, modern oil and gas drilling rigs have become ergonomic and tailored to the driller’s need for a safe and adequate working space. Modern drilling rigs rely on integrated mechanized equipment and ergonomic working space. Figure 8 shows an example of a highly computerized control system used to operate a modern rig, as opposed to the high level of labor required for a conventional rig.

Figure 8: Driller’s cabin on a modern drilling rig (Reinicke, 2005) compared to a driller’s working location in 1942.

Moreover, the paper “Increasing Geothermal Energy Demand: the Need for Urbanization of the Drilling Industry”, Bulletin of Science, Technology & Society 2008, Volume 28, No. 3, Special Issue, Part 2: The Age of Alternative Energy: Is the Future Renewable?, DOI:10.1177/0270467608315531, 16 April 2008.” shows the new trends in rig development and presents the future rig design from the authors point of view. To accommodate the future needs of drilling in urban environments, the new breed of drilling rigs will have the following design characteristics:

Driller

22

• Modular construction for easier mobilization and transportation through narrow city streets. Being modular will mean that most of the rig equipment will be mounted in small containers of small size, which has the added benefit of reducing noise emissions by sound insulation of the containers’ walls.

• Low noise emissions outside the drill site. One of the mandatory changes to rig design to reduce noise levels is to lower the rig’s height by using single drilling rigs instead of triple rigs. Noise reduction is also obtained by using non-conventional hoist systems (e.g. rack and pinion system, linear hydraulic motors)

• High hook load capacity, despite the small size of the drilling rig, which can be achieved by means of non-conventional technologies.

• Easily camouflage for long term drilling projects.

• Smart power supply using on-site diesel generators or tapping into the city electrical power grid, and having low emissions of noise, vibrations and exhaust gases.

• Closed loop spill and mud processing unit to avoid any contamination of the rig site.

• Small size of drilling rig, which will be achieved through integrating the automated processes and computer controlled rig floor equipment.

3.7. Multiphase flow with sand transport in horizontal pipes

All the above presented work requires a good understanding of the conditions existing downhole. This can be done by experimental and theoretical studies which sometimes are not directly linked to drilling technology. This explains the rather wide subjects in the selected papers and reports. Another interesting subject in the oil industry is the multiphase flow aspects of drilling and production fluids.

A crucial point, still to be established in the prediction of oil-gas-sand multiphase production and transfer system performance, is the identification of relevant mechanisms describing sand particle transport. To resolve this issue, experimental investigations are made on the behavior of suspended sand particles in simulated oil-gas-sand multiphase pipe flows paying attention to the time-averaged local and global sand velocity and holdup. Simultaneous measurements of the time-averaged local and global sand velocity are made by digital imaging technique for better understanding of the oil-gas-sand multiphase flow hydrodynamics and sand transport mechanisms. The results show that flow regimes of the multiphase flows significantly influence sand transport in the pipe flow. The shape of the local and global sand particle velocity and holdup profiles are also strongly modified by flow regimes. Furthermore, the experimental results indicate that the transport effect of the

23

suspended sand particles can be enhanced by operating the multiphase flows under slug flow conditions. It is concluded that a new mechanism based on bubble-particle interaction needs to be considered in the modeling of sand transport behavior during oil-gas-sand multiphase production and transfer operations. This work is based on extensive experimental investigations that were performed on a state-of-the-art testing facility. Figure 9 shows the schematic of the experimental facility. The developed testing facility can be successfully used for simulation of drilling and production well conditions. This facility will also be used as part of the research effort to improve the cuttings transport in geothermal wells as a part of the 2009 started project gebo (Forschungsverbund gebo – Geothermie und Bohrtechnik).

Particle Holdup Profiles in Horizontal Gas-Liquid-Solid Multiphase Flow Pipeline, Chemical Engineering &Technology, Volume 28, No. 12, December 2005, ISSN 0930-7516.

3.8. Development of advanced testing facilities for multiphase flow investigations

This paper is an example of the author’s efforts to perform fundamental research on “near praxis” environment. The need of such facility was shown through an SPE Journal published paper (Multiphase Flow Modelling Based on Experimental Testing: A Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments, (original SPE 110116) SPE Projects, Facilities & Construction, Volume 3, Number 3, September 2008, pp.1-10), that dealt with the current testing facilities worldwide and their applicability to specific research outcomes. Once the niche had been identified, a team has been built and the results are shown as follows.

Existing models to predict and analyze liquid loading in gas wells are based on steady-state flow. Even when transient multiphase wellbore models are employed, steady-state or pseudo steady-state inflow performance relationships are used to characterize the reservoir. A more reliable approach consists of modeling the dynamics in the near-wellbore region with its transient boundary conditions for the wellbore. The development of new models to mimic these dynamics requires a purpose-built flow loop. A design to construct such a facility has been developed.

This facility is the first to integrate pipe representing the wellbore with a porous medium that fully mimics the formation surrounding the wellbore. This design accounts not only for flow into the wellbore, but for any reverse flow from the pipe into the medium. Integrated wellbore/reservoir system analysis was used to screen the parameters required to recreate liquid loading under laboratory conditions. Once the range in operating conditions was defined, the equipment and mechanical components for the facility were selected and designed. The results showed that three reciprocating compressors working in parallel provide the smallest, most economic, and most flexible configuration for the Tower Lab facility at Texas A&M

24

University. The design of the pressure vessel hosting the porous medium requires a cylindrical body with top- and bottom-welded flathead covers with multiple openings to minimize weight. The required superficial velocities for air and water indicate that the system needs independent injection into the porous medium through two manifolds. Optimally, the system uses digital pressure gauges, coriolis or vortex technology to measure airflow and turbine meters for water flow. The new facility significantly improves the ability to mimic the physics of multiple phase flow for the development of liquid loading models and leads to better optimization of gas fields.

Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells, SPE-122786-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference to be held 31-May-09 to 03-Jun-09 in Cartagena, Colombia.

Figure 9: Schematic diagram of the experimental setup as shown in the paper: Particle Holdup Profiles in Horizontal Gas-Liquid-Solid Multiphase Flow Pipeline, Chemical Engineering &Technology, Voume 28, No. 12, December 2005, ISSN 0930-7516.

3.9. Close view on the low recovery factor in many oil fields and discussions of possible improvements

Technological developments and the decision to use high technology (HighTech) strongly depend on the reserves and resources which are to be found. This is why, drilling and production operation researchers have to look in the aspects of recovery factors and definition of reserves. This will enhance the capacity of generating applicable research, thus the fundamental aspect is preserved.

25

Accepted wisdom suggests that the higher an oil or gas fields’ value of recovery factor (RF), the more efficient the hydrocarbons have been produced from the reservoir. Optimizing the recovery from a hydrocarbon field should be the common goal of both governments and operators, although increasing production levels at the right economic and political moment may prove too tempting to some. Hence, the use of RF as a yardstick to measure the performance of a reservoir (or the management of that reservoir) has some serious shortcomings. In order to use RF appropriately, it is important to understand what it is, how it is calculated and the uncertainty inherent to the parameters from which it is derived. The value of RF is defined as the ratio of the recoverable volume to the hydrocarbons originally in place (HOIP) over the course of a field’s economic life. Yet this seemingly trivial calculation has inherent uncertainty and can vary due to many reasons.

Table 1: Template for what production data should be provided by government agencies. [WHP=Well Head Pressure; WHT=Well Head Temperature; GOR=Gas-Oil Ratio; WOR=Water-Oil Ratio; EOR=Enhanced Oil Recovery; PVT=Pressure-Volume- Temperature; FVF=Formation Volume Factor; CGR=Condensate-Gas Ratio; STOIIP=Stock Tank Oil Initially In Place]

This paper critically reviews the concept of the recovery factor of oil and gas fields. Although this simple parameter is used throughout the oil and gas industry, it is subject to misunderstanding and misuse. Besides changing continually through the producing life of a field, the estimate of RF is affected by geological uncertainty, inappropriate reserves reporting, technological shortcomings, commercial practices and political decisions. At present, the information necessary to fully evaluate RF is not unequivocally determined, audited or reported, which makes it impossible to produce consistent global field statistics. Based on the authors’ experience, the paper outlines the shortcomings of RF and suggests how they may be overcome. To

Item to be Reported Reported Variables for Item Outputs Derived from Variables

1a Monthly well producer recordsgas-oil-water volumes, days online, WHP, WHT, choke %

Producer Uptime, Well producing GOR-WOR-Water Cut, Well Decline Curve, Reserves per well, Field Reserves

1b Monthly well injector records gas, water volumes, days online, WHP, WHTInjector Uptime, Injectivity model, Producer:Injector Ratio

1c Well Technology

Artificially lifted? Stimulated? Water shut-offs? Horizontal or vertical? Completion details Aids selection of appropriate analogue fields

3a Monthly Field production records gas-oil-water volumesField producing GOR-WOR-Water Cut, Field Decline Curve, Field Reserves

3b Monthly Field injection records gas, water volumes Field Injectivity model (aquifer strength)

3c Field TechnologyGas compression? Multiphase pumping? EOR? Aids selection of appropriate analogue fields

4 Bottom hole Surveysaverage reservoir pressure & temperature history (date and depth datum) Material balance for HOIP

5 Top & Base reservoir structure mapsdepth contours, scale, well locations, fluid contacts, major faults

Gross rock volume, Field area, Hydrocarbon fill factor

6 Field Geological descriptionsandstone or carbonate, matrix porosity or naturally fractured, massive or thin-bedded Aids selection of appropriate analogue fields

7 Field Petrophysical parametersporosity, water saturation, gross thickness, net-to-gross

Hydrocarbon pore volume (at reservoir conditions) [Combining Items 5 & 7]

8 Field PVT properties (Oil)API gravity, solution GOR, viscosity, bubble point, FVF

STOIIP (at surface conditions) [Combining Items 5, 7 & 8]

9 Field PVT properties (Gas)Gas gravity, condensate gravity, CGR, viscosity, dew point, H2S-CO2-N2, FVF

GIIP (at reservoir conditions) [Combining Items 5, 7 & 9]

10 Recovery FactorField Reserves (from 1a or 3a), HOIP (from 8 or 9), with date reference

26

promote clarity and transparency in RF calculations, a template for an open worldwide production database is proposed. Table 1 shows the minimum structure of such data base.

Can We Be More Efficient in Oil and Gas Exploitation? A Review of the Shortcomings of Recovery Factor and the Need for an Open Worldwide Production Database, Scientific Journals International, Journal of Physical and Natural Sciences, Volume 1, Issue 2, 2007.

27

References Aloui, F. and Souhar, M. (1996): Experimental study of a two-phase bubbly flow in a

flat duct symmetric sudden expansion — Part II: liquid and bubble velocities, bubble sizes, International Journal of Multiphase Flow, Vol. 22, No. 5, 849-861.

Appah, D. and Ichara, M. 1994 Empirical model determines energy required to clean

Sand from wellbores, Oil and Gas Journal, Feb. 1994, 36-38. Appah, D.; Ichara, M. and Bouhroun, A. 1997 Aerated wash over technique in sand

producing wells, Oil and Gas European Magazine, Vol. 4, 1997, 29-33. Bello, O. O. and Fasesan, S. O. 2003 Improved Profitability in Nigerian Hydrocarbon

Production through Application of Sand Management Technology, Proceedings of the 33rd Annual Conference and Meeting of Nigerian Society of Chemical Engineers, Abeokuta, Nigeria, 19th to 21st Nov. 2003.

Bello, O. O. and Fasesan, S. O. 2003 Sand Management Technology for Improved

Oil Recovery in Nigeria” NAFTA Oil and Gas Scientific Journal, Year 55, Number 7-8, July-August, 2003.

Birger, I. A. 1944, Spannungsverteilung auf die Gewindeanzeiger des Maschinen-

baues, Moskau, Heft 11. Bratli, R. K.; Dusseault, M. B.; Santarelli, F. J. and Tronvoll, J. 2000, Sand

Management Protocol Increases Production Rate, Reduces Completion Costs, Proc. Trinidad and Tobago Biennial SPE Conf., Port-of-Spain.

Caicedo, G. R.; Marques, J. J.; Ruiz, M. G. and Soler, J. G. 2002, A study on the

behaviour of bubbles of a 2D gas-liquid fluidized bed using digital image analysis, Chemical Engineering and Processing, Vol. 42, 2002, 9-14.

Dusseault, M. B. and El-Sayed, S. 2001, Heavy Oil Production Enhancement by

Encouraging Sand Production, Proceedings 2001 SPE/DOE Improved Oil Recovery Symposium, Tulsa, Oklahoma, 3-5 April 2001.

Dusseault, M. B.; Geilikman, M. B. and Spanos, T. J. T. 1998 Mechanism of Massive

Sand Production in Heavy Oils, Proc. 7th UNITAR Int. Conf. Heavy Oils and Tar Sands, Beijing.

Dusseault, M. B.; Tronvolli, J.; Santilippo, F and Santarelli, F. J. 2000, Sand-Self

Cleaning in High Rate Oil Wells Using Sand Management, SPE 58786, Int. Conf. on Formation Damage.

Dusseault, M. B.; Chun, X. L.; Yiqiu, M. and Wu, G. 2002, CHOPS in Jilin Province,

China, Proceeding of SPE International Horizontal Well Technology Conf., Calgary, Alberta, Canada, 4-7 November 2002.

Ehsan, D.; Kegang, L.; Teodoriu, C.; William, D.; McCain, Jr.; and Falcone, G. 2009,

More Accurate Gas Viscosity Correlation for Use at HPHT Conditions Ensures

28

Better Reserves Estimation, SPE 124734, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

Falcone, G.; Hewitt, G. F.; Alimonti, C. 2009, Multiphase Flow Metering: Principles

and Applications, Elsevier, Textbook for the Developments of Petroleum Science series. Elsevier, October 2009.

Geilikman, M. B. and Dusseault, M. B. 1997, Fluid-rate Enhancement from Massive

and Production in Heavy oil Reservoirs, J. of Petr. Science and Engineering, 17, 5-8. Special Issue: Near Well bore Formation Damage and Remediation, 1997.

Gillies, R. G.; Hill, K. B.; Mckibben, M. J. and Shook, C. 1999, Solids transport by

laminar Newtonian flows, Powder Technology, Vol. 104, 1999, 269-277. Gillies, R. G.; Mckibben, M. J. and Shook, C. 1997, Pipeline flow of gas, liquid and

sand mixture at low velocity, J. Can. Petro. Tech. 36, 1997, 36-42. Gillies, R. G.; Mckibben, M. J. and Shook, C. 1995, Oil, water and sand flow

experiments in a model horizontal well, J. Can. Petro. Tech. 34, No. 9, 1995, 56-63.

Gopal, M. and Jepson, W.P. 1997, Development of digital image analysis techniques

for the study of velocity and void profiles in slug flow, International Journal of Multiphase Flow, Vol. 23, No. 5, 1997, 945-965.

Jepson, W. P. 1987, The flow characteristics in horizontal slug flow, Proc. 3rd Int.

Conf. on Multiphase Flow, B.H.R. Group, The Hague, Netherlands 18-20 May, 1987.

Kaftori, D.; Hetstroni, G. and Banerjee, S. 1995, Particle behaviour in the turbulent

layer: motion, deposition and entrainment, Physics of Fluid, Vol. 7, No. 5, 1995,1095-1106.

Kaftori, D.; Hetstroni, G. and Banerjee, S. 1995, Particle behaviour in the turbulent

layer: velocity, flux and concentration distribution, Physics of Fluid, Vol. 7, No. 5, 1995, 1107-1121.

King, M. J. J.; Fairhurst, C. P. and Hill, T. J. 2001, Solids transport in multiphase

flows: applications to high viscosity systems, Journal of Energy Resources, Vol. 123, 2001. 200-204.

Kundakovic, L. and Vunjak-Novaakovie, G. 1995, Mechanics of Particle Motion in

Three-phase flow, Chemical Engineering Science, Vol. 50, No. 20, 1995, 3285-3295.

Leifer, I.; Leeuw, G-d.; Kunz, G. and Cohen, L. H. 2003, Calibrating optimal bubble

size by the displaced-mass method, Chemical Engineering Science, Vol. 58, 2003, 5211-5216.

29

Marx, C. Teodoriu, C., Pupazescu, Al., 2001, A New Method to Determine the Stress Distribution in Conical Shouldering Thread Connections, DGMK-Frühjahrstagung, 2001.

Oudeman, P. 1993, Sand Transport and Deposition in Horizontal Multiphase Trunk

lines of Sub-sea Satellite Developments, SPE 25142. Sager, J. A 1986, Contribution to the Determination of the Stress Distribution in

Shouldered Threaded Connections under Axial Load, ITE. Sakaguchi, T.; Minagawa, H. and Tomiyama, A. 1997, Slug characteristics and slug

modelling of the gas-liquid-solid three-phase slug flow in vertical pipes, Proceedings of the Experimental Heat Transfer, Fluid mechanics and Thermodynamics, Giot, M., Mayinger, F. and Celata, G. P. (editors), 1997, 1137-1144.

Sakaguchi, T.; Minagawa, H.; Sahara, K.; Kato, Y.; Kuroda, N. and Matsumoto, T.

1987, Estimation of volumetric fraction of each phase in gas-liquid-solid three-phase flow, Proceedings of the JSME/ASME Thermal Engineering Joint International Conference, Honolulu, Hawaii, March, 1987.

Sakaguchi, T.; Minagawa, H.; Saibe, T. and Sahara, K. 1988, Estimation of

volumetric fraction of each phase in gas-liquid-solid three-phase slug flow, Proceedings of the Japan-US Seminar on Two-phase Flow Dynamics, Ohtsu, Japan, July, 1988.

Sakaguchi, T.; Minagawa, H.; Tomiyama, A. and Shakutshi, H. 1992, Gas-liquid-solid

three-phase flow in a vertical pipe, International Video Journal of Engineering Research, Vol. 2, 1992, 37-45.

Sakaguchi, T.; Shakutsui, H.; Tomiyama, A.; Minagawa, H. and Takahashi, H. 1991,

Flow characteristics of gas-liquid-solid three-phase bubbly flow in vertical pipes, proceedings of the 1st JSME/ASME Joint International Conference on Nuclear Engineering, Tokyo, Japan, November, 1991.

Salama, M. M. 1998, Sand Production Management”, OTC 8900, Houston, 4-7 May

1998. Shen, I.; Johnson, F. and Leckner, B. 2003, Digital image analysis of hydrodynamic

two-dimensional bubbling fluidized beds, Chemical Engineering Science, Vol. 59, 2003, 2607-2617.

Stevenson, P. 2001, Particle transport in pipes by two-phase flows, Ph.D. Thesis, University of Cambridge, Cambridge, United Kingdom. Stevenson, P. and Thorpe, R. B. 1999, Towards understanding sand transport in oil

flowlines, Proc. 9th Int. Conf. on Multiphase Flow, B.H.R. Group, Cannes,France 16-18 June, 1999.

Stevenson, P. and Thorpe, R. B. 2002, Method calculates sand velocity, hold-up in

flowlines, Oil and Gas Journal, 100 (30), 47-50.

30

Stevenson, P. and Thorpe, R. B. 2003, Energy dissipation at the slug nose and the

modeling of solids transport in intermittent flow, Can. J. Chem. Engineering, 81, 271- 278.

Stevenson, P., Kennedy, J.E. and Thorpe, R. B. Prediction of sand behaviour in

intermittent flow using the long slug approximation, Proc. 12th Int. Conf. on Multiphase Technology, B.H.R. Group, Bantif, Canada, 2002

Stevenson, P.; Thorpe, R. B.; Kennedy, J. E. and McDermott, C. 2001, The transport

of particles at low loading in near-horizontal pipes by intermittent flow, Chemical Engineering Science, 56, 2149-2159.

Stevenson, P.; Thorpe, R. B.; Kennedy, J. E. and McDermott, C. 2001, The similarity

of sand transport by slug flow and hydraulic conveying, Proc. 10th Int. Conf. on Multiphase Flow, B.H.R. Group, Cannes, France 13-15 June 2001.

Teodoriu, C. 2003, Analysis of the make-up procedure and evaluation of conical

shouldered connections, Ph.D. Thesis, Institute of Petroleum Engineering, Technical University Clausthal.

Tronvoll, J.; Dusseault, M. ; Sanfilippo, F. and Santarelli, F. J. 2001, The Tools of

Sand Management, Annual Tech. Conf. and Exhibition, New Orleans, Louisiana, 30 Sept. -3 Oct. 2001.

Wicks, M. 1971, Advances in solid-liquid flow in pipes and its application. Ed. Zandi,

I. Pergamon Press, 101-124. Ulmanu, V. 1973, Ph.D. Thesis, Universitatea Petrol-Gaze, Ploiesti. ***, API RECOMMENDED PRACTICE 7A1 (RP 7A1) Recommended practice for

testing of thread compound for rotary connections, American Petroleum Institute, first edition , 1992

***, API RECOMMENDED PRACTICE 7G (RP 7G), 1992

31

Appendix 1: List of papers and other publications

32

Chapters in Books

1. Delgado, J. C.; Schubert, J. J.; Teodoriu, C.

OFFSHORE DRILLING AND PRODUCTION, prepared for ENCYCLOPEDIA OF LIFE SUPPORT SYSTEMS (EOLSS) by UNESCO, online since 2010.

Journal Papers

2. Teodoriu, C.; Reichetseder, P.; Marx, C.; Kinzel, H.

An introduction to Intelligent Make-up facilities: Analysis of Torque-Turn Recordings to make-up Rotary-Shouldered-Connections (RSC), Oil and Gas Magazine, 4/2002, ISSN 0342-5622.

3. Ulmanu, V.; Teodoriu, C.

Fatigue Life prediction and Test Results of Casing Threaded Connection, Mecanica Ruperii, Buletinul ARMR, No. 17, July 2005, ISSN14538148.

4. Reinicke, K. M.; Teodoriu, C.

Neue Entwicklungen in der Bohrtechnik, Akademie der Geowissenschaften zu Hannover Veröffentlichungen, Heft 25, 2005, ISBN 3-510-95943-4.

5. Teodoriu, C. Buttress Connection Resistance under Extreme Axial Com-pression Loads, Oil and Gas Magazine, 4/2005, Volume 31, ISSN 0342-5622.

6. Bello, O. O.; Reinicke, K. M.; Teodoriu, C.

Particle Holdup Profiles in Horizontal Gas-Liquid-Solid Multi-phase Flow Pipeline, Chemical Engineering &Technology, Vol 28, No. 12, November 2005, ISSN 0930-7516.

7. Falcone, G.; Harrison, B.; Teodoriu C.

Can We Be More Efficient in Oil and Gas Exploitation? A Review of the Shortcomings of Recovery Factor and the Need for an Open Worldwide Production Database, Scientific Journals International, Journal of Physical and Natural Sciences, Volume 1, Issue 2, 2007.

8. Falcone, G.; Teodoriu, C.; Reinicke, K. M.; Bello, O. O.

Multiphase Flow Modeling Based on Experimental Testing: A Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments, (original SPE 110116) SPE Projects, Facilities & Construction, Volume 3, Number 3, September 2008, pp.1-10.

9. Teodoriu, C.; Falcone, G.

Increasing Geothermal Energy Demand: the Need for Urbanization of the Drilling Industry, Bulletin of Science, Technology & Society 2008, volume 28, No. 3, Special Issue, Part 2: The Age of Alternative Energy: Is the Future Renewable?, DOI:10.1177/0270467608315531, 16 April 2008.

10. Surendra, M.; Falcone, G.; Teodoriu, C.

Investigation of Swirl Flows Applied to the Oil and Gas Industry, (original SPE 115938), accepted in August 2008 for publication in the SPE Projects, Facilities & Construction Journal.

11. Teodoriu, C.; Comparing completion design in hydrocarbon and geothermal

33

Falcone, G. wells: the need to evaluate the integrity of casing connections subject to thermal stresses, Geothermics journal, doi:10.1016/ j.geothermics.2008.11.006, 9 January 2009.

12. Teodoriu, C.; Badicioiu, M.

Sealing Capacity of API Connections - Theoretical and Experi-mental Results, SPE 106849, SPE Drilling and Completion Journal, March 2009.

13. Ulmanu, V.; Teodoriu, C.; Marx, C.; Pupazescu, A.

Untersuchung zur Belastbarkeit von SLIMHOLE-Bohrgestänge mit Doppelschulter, Erdöl, Erdgas, Kohle, Heft X, 125 Jg. 2009.

14. Park, H. Y.; Falcone, G.; Teodoriu, C.

Decision matrix for liquid loading in gas wells for cost/benefit analyses of lifting options, Journal of Natural Gas Science and Engineering, June 2008.

15. Teodoriu, C. Rotary-Shouldered Connections Make-up Torque Calculation Considering the Effect of Contact Pressure on Thread Compound’s Friction Coefficient; OIL GAS European Magazine 4/2009.

16. Falcone, G.; Chava, G.; Teodoriu. C.

“Plunger Lift Modeling Toward Efficient Liquid Unloading in Gas Wells”, SPE Projects, Facilities & Construction Journal, March 2010.

17. Fernandez, J.; Falcone, G.; Teodoriu, C.

“Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells”, SPE Projects, Facilities & Construction Journal, June 2010.

18. Zhang, H.; Falcone, G.; Teodoriu, C.

“Modeling Fully-Transient Two-phase Flow in the Near-Wellbore Region During Liquid Loading in Gas Wells”, accepted in April 2010 for publication in the Journal of Natural Gas Science & Engineering.

19. Limpasurat, A.; Falcone, G.; Teodoriu, C.; Barrufet, M. A.; Bello, O. O.

“Artificial Geothermal Energy Potential of Steam-Flooded Heavy Oil Reservoirs”, accepted in February 2010 for publication in the Int. J. Oil, Gas and Coal Technology.

20. Davani, E.; Ling, K.; Teodoriu, C.; McCain, W. D.; Falcone, G.

“Accurate Gas-Viscosity Correlation for Use at HP/HT Conditions Ensures Better Reserves Estimation”, Journal of Petroleum Technology (JPT), April 2010.

34

Full Peer-reviewed Conference Papers

21. Bello, O.; Reinicke, K. M.; Teodoriu, C.

Measurement of sand velocity and hold-up distributions in multiphase pipeline flow using digital image analysis techniques, Proceeding of the BHRl Multiphase Production Technology Conference 2005, Barcelona, Spain, Volume 1, May 25-27, 2005, pp. 171-189.

22. Bello, O. O.; Reinicke, K. M.; Teodoriu, C.

Experimental Study on Particle Behavior in a Simulated Gas-Oil-Sand Multiphase Production and Transfer Operation. Proceedings of the 2006 ASME Joint US - European Fluids Engineering Division Summer Meeting and Exhibition, FEDSM2006-98353, Miami, FL, USA, July 17-20 2006.

23. Teodoriu, C.; Ohla, K.; Nielsen, W.

The future of environmental riserless drilling: dope free drill pipe connections using thread saver technology, OMAE-29459, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

24. Teodoriu, C.; Kinzel, H.; Schubert, J.

Evaluation of Real Time Torque-Turn Charts with Application to Intelligent Make-up Solutions, OMAE2007-29518, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

25. Teodoriu, C.; McDonald, H.; Bolfrass, C.

Friction Considerations in Rotary Shouldered Threaded Connections, OMAE2007-29583, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

26. Bello, O. O.; Falcone, G.; Teodoriu, C.

Experimental Validation of Multiphase Flow Models and Testing of Multiphase Flow Meters: a Critical Review of Flow Loops Worldwide, International Multiphase Flow Conference and Exhibition 2007, Bologna, 2007.

27. Teodoriu, C.; Falcone, G.; Espinel, A.

Letting Off Steam and Getting Into Hot Water – Harnessing the Geothermal Energy Potential of Heavy Oil Reservoirs, World Energy Congress 2007, ROME 2007.

28. Solomon, F.; Falcone, G.; Teodoriu, C.

The Need to Understand the Dynamic Interaction Between Wellbore and Reservoir in Liquid Loaded Gas Wells, 27th International Conference on Offshore Mechanics and Arctic Engineering (OMAE 2008), Estoril, Portugal, 15-20 June 2008.

29. Teodoriu, C.; Falcone, G.

Investigation of Drilling Problems in Gas Hydrate Formations, 27th International Conference on Offshore Mechanics and Arctic Engineering (OMAE 2008), Estoril, Portugal, 15-20 June 2008.

35

Conferences

30. Teodoriu, C.; Dinulescu, V.

Study of Casing Loads in Steam Stimulated Wells, ICPT Campina, 1997.

31. Teodoriu, C.; Dinulescu, V.

A Discussion about Casing Performance Under Thermal Cycling Conditions, The first Huf’n’Puff Project in Romania, ICPT Campina, 1997.

32. Ene, C. D.; Teodoriu, C.; Arvunescu, M.

About Kinematics of Reciprocal Pumps without Pulsation, Bul. Universitatii „Petrol-Gaze“ Ploiesti, Vol. XLVII, Nr. 6, 1998.

33. Popovici, A.; Ene, C. D.; Olteanu, M.; Teodoriu, C.

Developing a Dynamical Model of a Sonic Pump Unit, Bul. Universitatii „Petrol-Gaze“ Ploiesti, Vol. XLVII, Nr. 6, 1998.

34. Marx, C.; Teodoriu, C.; Pupazescu, A.

A new method to determine the stress distribution in conical shouldering thread connections, DGMK, Frühjahrstagung, 22./23. April 2002, Band 1, ISBN 3-931850-90-0, ISSN 1433-9013.

35. Teodoriu, C.; Reichetseder, P.

A review of analytical calculation of shouldered connections, DGMK, Frühjahrstagung, 28./29. April 2003, Band 1, ISBN 3-936418-03-9, ISSN 1433-9013.

36. Teodoriu, C.; Reichetseder, P.; Marx, C.; Kinzel, H.

Analysis of Torque-Turn Recordings to make-up Rotary-Shouldered-Connections (RSC), DGMK, Frühjahrstagung, 28./29. April 2003, Band 1, ISBN 3-936418-03-9, ISSN 1433-9013.

37. Popa, A.; Teodoriu, C.; Reichesteder, P.

Data management for small research network, SPC-2003, ”Petroleum-Gas” University of Ploiesti, 8th- 9th October 2003.

38. Teodoriu, C.; Popa, A.; Reichetseder, P.

Data Acquisition system for real-time strain gauge, measurement using DASYLab, SPC-2003,”Petroleum-Gas” University of Ploiesti, 8th- 9th October 2003.

39. Teodoriu, C. Contact pressure distribution on the thread flank of shouldered threaded connections using finite element analysis, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 29./30. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

40. Teodoriu, C.; Marx, C.; Reichetseder, P.

Analytical method to determine stress distribution in rotary shouldered connections, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 29./30. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

36

41. Ghofrani, R.; Teodoriu, C.; Stekolschikov, K.

Swelling cement induced forces and experimental swelling pressure, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 29./30. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

42. Pisarski, P.; Reichetseder, P.; Teodoriu, C.; Lieser, D.

Die Simulation des nicht-isothermen Gasdurchflusses in Pipe-lines mit der Berücksichtigung des Wärmespeichereffekts der Rohrumgebung, DGMK Tagungsbericht 2004-2, Frühjahrsta-gung, 29./30. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

43. Ulmanu, V.; Teodoriu, C.

Experimental Stress Analysis on a Double Shoulder Connection, 10th International Symposium on experimental Stress Analysis and Material Testing, Sibiu, Romania, 22-23 OCTOBER 2004.

44. Teodoriu, C.; Kinzel, H.

The Application of an Analytical model for the Controlled Make-up of Rotary Shouldered Connections in the Field, SPE Asia-Pacific Conference 2005, SPE 93777.

45. Bello, O. O.; Reinicke, K. M.; Teodoriu C.

Evaluation of the modes of transporting sand under conditions analogous to oil-gas multiphase pipeline system, DGMK Tagungsbericht 2005-1, Frühjahrstagung, 29./30. April 2005, ISBN 3-936418-17-9, ISSN 1433-9013.

46. Popa, A.; Teodoriu, C.

Sistem de ceasuri dinamice pentru cronometrarea experiment-telor multiple, SPC-2005,”Petroleum-Gas” University of Ploiesti, 11th- 13th May 2005.

47. Teodoriu, C.; Popa, A.; Reinicke, K. M.; Rothfuss, M.

Implementation of an online monitoring system for the ITE Center for testing oilfield country tubular goods, SPC-2005, ”Petroleum-Gas” University of Ploiesti, 11th- 13th May 2005.

48. Teodoriu, C.; Popa, A.; Klaws, M.

Optical system for leak measurement of PREMIUM threaded connections, SPC-2006,”Petroleum-Gas” University of Ploiesti, 17th- 19th May 2006.

49. Teodoriu, C.; Patil, P.

Evaluation of Surface Casing Resistance having Corrosion Damage, DGMK Tagungsbericht 2006-1, Frühjahrstagung, 29./30. April 2006, ISBN 3-936418-48-9, ISSN 1433-9013.

50. Ulmanu, V.; Badicioiu, M.; Teodoriu, C.

Laboratory for Testing Full Scale Oil Country Tubular Goods Subjected to Complex Loads, UPG Bulletin, 2006.

51. Badicioiu, M.; Teodoriu, C.

Sealing Capacity of API Connections - Theoretical and Experimental Results, SPE 106849, 2007 SPE Production and Operations Symposium held in Oklahoma City, Oklahoma, U.S.A., 31 March – 3 April 2007.

37

52. Teodoriu, C.; Schubert, J.

Redefining the OCTG Fatigue – A Theoretical Approach, 2007 Offshore Technology Conference held in Houston, Texas, U.S.A., 30 April – 3 May 2007.

53. Teodoriu, C.; Falcone, G.

How real-time monitoring technology will change the future of laboratory classes, Teaching with Technology Conference 2007, Held at Texas A&M University, College Station, 2007.

54. Falcone, G.; Teodoriu, C.; Reinicke, K. M.; Bello, O. O.

Multiphase Flow Modeling Based on Experimental Testing: a Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments, ATCE 2007.

55. Teodoriu, C.; Reinicke, K. M.; Holzmann, J.; Klaws, M.

Testing of Tubular Goods – a Critical Review of Past and Actual Testing Procedures, DGMK Tagungsbericht 2005-1, Frühjahrs-tagung, 28./29. April 2007.

56. Teodoriu, C.; Falcone, G.

Low cost completion for steam injections wells: theoretical and experimental results, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14-17 October 2007.

57. Teodoriu, C.; Ulmanu, V.; Badicioiu, M.

Casing Fatigue Life Prediction using Local Stress Concept: Theoretical and Experimental Results, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14-17 October 2007.

58. Haghshenas, A.; Amani, M.; Teodoriu, C.

Feasibility Study of Steam Injection in one of Iranian Naturally Fractured Heavy Oil Fields, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14-17 October 2007.

59. Paknejad, A.; Schubert, J.; Teodoriu, C.; Amani, M.

Sensitivity Analysis of Key Parameters in Foam Drilling Operations, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14-17 October 2007.

60. Teodoriu C.; Falcone, G.

Fatigue Life Prediction of Buttress Casing Connection Exposed to Large Temperature Variations, 33rd Stanford Workshop on Geothermal Reservoir Engineering, Fisher Conference Center on the Stanford University campus, 28-30 January 2008.

61. Teodoriu, C.; Falcone, G.

Comparison of Well Completions used in Oil/Gas Wells and Geothermal Wells: a New Approach to Technology Transfer, 33rd Stanford Workshop on Geothermal Reservoir Engineering, 28-30 January 2008, Fisher Conference Center on the Stanford University campus.

38

62. Teodoriu, C.; Ulmanu, V.; Badicioiu, M.

Casing Fatigue Life Prediction Using Local Stress Concept: Theoretical and Experimental Results, SPE 110785, 2008 SPE Western Regional Meeting held 31-MAR to 02-APR-08 in Los Angeles, CA., 2008.

63. Teodoriu, C.; Falcone, G.

Low-Cost Completion for Steam Injections Wells: Theoretical and Experimental Results, SPE 110454, 2008 SPE Improved Oil Recovery Symposium to be held 21 to 23 APR 2008 in Tulsa, OK., 2008.

64. Teodoriu C.; Falcone, G.

Oil and Gas expertise for Geothermal Exploitation: the Need for Technology Transfer", SPE 113852, SPE EUROPEC Con-ference, ROME 2007.

65. Ibeh, C.; Schubert, J.; Teodoriu, C.

Methodology for Testing Drilling Fluids under Extreme HP/HT Conditions, AADE 2008 conference held April 8-9, 2008 at the Hilton (formerly Wyndham) Greenspoint Hotel in Houston, Texas, 2008.

66. Ibeh, C.; Teodoriu, C.; Schubert, J.; Gusler, W.; Harvey, F.

Investigation on the Effects of Ultra-High Pressure and Tempera-ture on the Rheological Properties of Oil-Based Drilling Fluids, AADE 2008 conference to be held April 8-9, 2008 at the Hilton (formerly Wyndham) Greenspoint Hotel in Houston, Texas, 2008.

67. Teodoriu, C.; Schubert, J.; Vivek, G,; Ibeh, C.

Investigations to Determine the Drilling Fluid Rheology Using Constant Shear Rate Conditions, presented at the SPE IADC conference, held in Berlin, Germany, June, 2008.

68. Chava, G.; Falcone, G.; Teodoriu, C.

Development of a New Plunger-Lift Model Using Smart Plunger Data, SPE 115934, presented at the 2008 SPE Annual Technical Conference and Exhibition, Denver, Colorado, U.S.A., 21-24 September 2008.

69. Surendra, M.; Falcone, G.; Teodoriu, C.

Investigation of Swirl Flows Applied to the Oil and Gas Industry, SPE 115938, presented at the 2008 SPE Annual Technical Conference and Exhibition, Denver, Colorado, U.S.A., 21-24 September 2008.

70. Solomon, F.; Falcone, G.; Teodoriu, C.

Critical Review of Existing Solutions to Predict and Model Liquid Loading in Gas Wells, SPE 115933, presented at the 2008 SPE Annual Technical Conference and Exhibition, Denver, Colorado, U.S.A., 21-24 September 2008.

39

71. Fernandez, J.; Falcone, G.; Teodoriu, C.

Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells, SPE-122786-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference to be held 31-May-09 to 03-Jun-09 in Cartagena, Colombia, 2009.

72. Zhang, H.; Falcone, G.; Valko, P.; Teodoriu, C.

Numerical Modeling of Fully-Transient Flow in the Near-Wellbore Region During Liquid Loading in Gas Wells, SPE-122785-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference to be held 31-May-09 to 03-Jun-09 in Cartagena, Colombia, 2009.

73. Ehsan, D.; Kegang, L.; Teodoriu, C.; McCain, W. D.; Falcone, G.

Inaccurate Gas Viscosity at HP/HT Conditions and its Effect on Unconventional Gas Reserves Estimation, SPE-122827-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference to be held 31-May-09 to 03-Jun-09 in Cartagena, Colombia, 2009.

74. Teodoriu, C.; Reinicke, K. M.; Falcone, G.

Real-Time Long Distance Teaching: An Overview of Two Years of Tele-Teaching between Texas and Germany, SPE 124748, paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

75. Ehsan, D.; Kegang, L.; Teodoriu, C.; McCain, W. D. Jr.; Falcone, G.

More Accurate Gas Viscosity Correlation for Use at HPHT Conditions Ensures Better Reserves Estimation, SPE 124734, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

76. Chava, G.; Falcone, G.; Teodoriu, C.

Plunger Lift Modeling Towards Efficient Liquid Unloading in Gas Wells, SPE 124515, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

77. Chava, G.; Falcone, G.; Teodoriu, C.

Use of Integrated Electronic and Sensors to Improve Plunger Lift Modeling Towards Efficient Liquid Unloading in Gas Wells, DGMK/ÖGEW-Frühjahrstagung 2009, 27./28. April 2009.

78. Teodoriu, C.; Falcone, G.; Fichter, C.

Drilling Beyond Oil and Gas: a Discussion about How Technology will Change to Enhance Drilling in Urban Areas, DGMK/ÖGEW-Frühjahrstagung 2009, 27./28. April 2009.

79. Ling, K.; Teodoriu, C.; Ehsan, D.; Falcone, G.

“Measurement of Gas Viscosity at High Pressures and High Temperatures”, International Petroleum Technology Conference, Doha, Qatar, 7–9 December 2009.

40

80. Reinicke, K. R.; Ostermeyer, G. P.; Overmeyer, L.; Teodoriu C.; Thomas, R.

Der Niedersächsische Forschungsverbund Geothermie und Hochleistungsbohrtechnik (gebo), Energie und Rohstoffe 2009, Goslar, 2009.

81. Popa, A.; Teodoriu, C.

Automatic Test Facility for the Wellbore Cement Integrity, 6th International Symposium on Process Control, SPC-2009, Ploiesti, June 1 – 3, 2009.

82. Popa, A.; Teodoriu, C.; Popa, C.

Analysis of the Nonlinearities in the Hydraulic Actuators 6th International Symposium on Process Control, SPC-2009, Ploiesti, June 1 – 3, 2009.

83. Fichter, C.; Reinicke, K. M.; Teodoriu, C.

Tiefbohrtechnik – Erschließung von tiefgeothermischen Wärme-speichern, GeoTHERM 2009, Offenburg, April 2009.

84. Teodoriu, C.; Fichter, C.; Falcone, G.

Drilling Deep Geothermal Reservoirs: the Future of Oil and Gas Business, Beitrag “Der Geothermiekongress 2009” Bochum, Germany, 17.-19. November 2009.

85. Fichter, C.; Reinicke, K. M.; Teodoriu, C.; Wang, Y.; Lotz, U.

Influences and limiting parameters of casing design on the success of hydrogeothermal projects - a new approach for the design of geothermal wells, Beitrag “Der Geothermiekongress 2009” Bochum, Germany, 17.-19. November 2009.

86. Teodoriu, C.; Reinicke, K. M.; Fichter, C.; Wehling, P.

Investigations on Casing-Cement Interaction with Application to Gas and CO2 Storage Wells, SPE 131336, EUROPEC/EAGE Annual Conference and Exhibition held in Barcelona, Spain, 14–17 June 2010.

87. Teodoriu, C.; Ugwu, I.; Schubert, J.

Estimation of Casing-Cement-Formation Interaction using a new Analytical Model, SPE131335, SPE EUROPEC/EAGE Annual Conference and Exhibition held in Barcelona, Spain, 14–17 June 2010.

88. Khabibullin,T.; Falcone, G.; Teodoriu, C.; Bello, O. O.

“Drilling through Gas Hydrate Sediments: Managing Wellbore Instability Risks”, SPE EUROPEC/EAGE Annual Conference and Exhibition held in Barcelona, Spain, 14–17 June 2010.

89. Teodoriu, C.; Fichter, C.

Anpassung der Tiefbohrtechnik aus der Erdöl-/Erdgasindustrie an die Tiefengeothermie, 3. Fachtagung, Geothermische Tech-nologien – Technologien zur Nutzung der tiefen Geothermie und ihre Integration in Energieversorgungssysteme, 23./24. März 2010.

90. Gedzius, I.; Teodoriu, C.; Fichter, C.; Reinicke, K. M.

Betrachtung von Korrosionsschäden an Casings im Primärkreislauf geothermischer Anlagen, DGMK/ÖGEW-Frühjahrstagung 2010, Fachbereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

91. Pilisi,N.; An Advisory System for Selecting Drilling Technologies and

41

Teodoriu, C. Methods with Application for Tight Gas Reservoirs, DGMK/ÖGEW-Frühjahrstagung 2010, Fachbereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

92. Bai, M.; Fichter, C.; Teodoriu, C.; Reinicke, K. M.

Development of a Novel Testing Method to Perform Investigations on Rock-Fluid Interaction under Geothermal Hot Dry Rock Conditions, DGMK/ÖGEW-Frühjahrstagung 2010, Fachbereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

93. Teodoriu, C.; Baruffet, M.; Epps, M.L.; Falcone, G.; Reinicke, K. M.

Real-Time Long Distance Teaching: The Texas A&M DL Experience, ATCE 2010, Florence, Italy, Sept. 2010.

94. Teodoriu, C. Hands-on Teaching as Part of Petroleum Engineers Education: An International Experience, ATCE 2010, Florence, Italy, Sept. 2010.

95. Reinicke, K. M.; Oppelt, J.; Ostermeyer, G. P.; Overmeyer, L.; Teodoriu, C.; Thomas, R.

Enhanced technology transfer for geothermal exploitation through a new research concept: The Geothermal Energy and High Performance Drilling Research Program – gebo, ATCE 2010, Florence, Italy, Sept. 2010.

96. Bello, O. O.; Falcone, G.; Teodoriu, C.; Udong, I.

“Hydraulic Analysis of Gas-Oil-Sand Flow in Horizontal Wells”, SPE 136874 to be presented at the SPE Latin American & Caribbean Petroleum Engineering Conference, Lima, Peru, 1–3 December 2010.

97. Zhang, H.; Falcone, G.; Valko, P.; Teodoriu, C.

"Relative Permeability Hysteresis Effects in the Near-Wellbore Region During Liquid Loading in Gas Wells", SPE 139062 to be presented at the SPE Latin American & Caribbean Petroleum Engineering Conference, Lima, Peru, 1–3 December 2010.

98. Limpasurat, A.; Falcone, G.; Teodoriu, C.; Barrufet, M.

“Unconventional Heavy-Oil Exploitation for Waste Energy Recovery”, SPE 139054 to be presented at the SPE Latin American & Caribbean Petroleum Engineering Conference, Lima, Peru, 1–3 December 2010.

Submitted papers, currently under review

99. Bello, O. O.; Falcone, G.; Teodoriu, C.; Udong, I.

“Optimizing Two-Phase Oil-Sand Flow through a Horizontal Well”, submitted to the Oil & Gas Science and Technology - Revue de l'IFP in December 2009.

42

100. Khabibullin,T.; Falcone, G.; Teodoriu, C.; Bello, O. O.

“Drilling through Gas Hydrate Sediments: Managing Wellbore Instability Risks”, submitted to the SPE Drilling & Completion journal in March 2010.

101. Limpasurat, A.; Falcone, G.; Teodoriu, C.; Barrufet, M. A.; Bello, O. O.

“Artificial Geothermal Energy Potential of Steam-Flooded Heavy Oil Reservoirs”, accepted in February 2010 for publication in the Int. J. Oil, Gas and Coal Technology.

Invited Speaker and Presentations to Industry Boards

102. Teodoriu, C. Formation damage and stimulation, held at the “Industry meets University“ Workshop, Institute of Petroleum Engineering, TU-Clausthal, Germany, 06.02.2004.

103. Teodoriu, C. Introduction to Intelligent Make-up Facilities, German Section of SPE, ITE, TU Clausthal, 25. June 2004.

104. Teodoriu, C. Tiefpumpen, eine Übersicht über Bauarten und Anwendungen, presentation at the Tiefpumpen Workshop, held in Karlsruhe, 06.07.2005.

105. Teodoriu, C. Use of Analytical Models for the Development of Intelligent Make-up Machines, held at Texas A&M, Department of Petroelum Engineering, J. L. "Corky" Frank '58 Graduate Seminar Series, 14 Nov. 2006.

106. Teodoriu, C.; Falcone, G.

The future of laboratory classes: virtual and remote controlled experiments using real-time monitoring and web integrated approaches, held at Texas A&M, Department of Petroleum Engineering, J. L. "Corky" Frank '58, Graduate Seminar Series, 13 Feb. 2007.

107. Teodoriu, C. Well Construction 2020+, SPE 2008 Forum Series, held in Dubrovnik, Croatia, invited as discussion leader for Cultural Changes and Human Resources Session, 14-19 September 2008.

108. Reinicke, K. M.; Ostermeyer, G. P.; Overmeyer, L.; Teodoriu C.; Thomas, R.

Der Niedersächsische Forschungsverbund Geothermie und Hochleistungsbohrtechnik (gebo), HotSpot Geothermietag, Hannover, 29.10.2009.

43

109. Teodoriu, C. Intelligent Oil Country Tubular Goods Make-up: An Alternative to Increase Tubular Reliability, BERG- UND HÜTTENMÄNNISCHER TAG, Freiberger Forschungsforum „Ressourcen für die Zukunft“ 17.-19.06.2009.

Industry Seminars and short-courses

1. Reichetseder, P.; Teodoriu, C.

Kurz-Seminar, Petroleum Production Engineering, at J. H. Bornemann GmbH, Obernkirchen, 9. Sept. 2003.

2. Teodoriu, C. Kurz-Seminar, Oil Production Systems, at BASF, Ludwigshafen, May 2006.

3. Teodoriu, C. Short Seminar, Drill String mechanics, Perforator GmbH, Walkenried, Germany, April 2006.

4. Teodoriu, C. Short Seminar, Drill String mechanics, Weatherford Oil Tools GmbH, Germany, April 2006.

5. Teodoriu, C. Stuck pipe course, on behalf of NExT Training Center for PETROM S.A., Ploiesti, Romania, June and Nov. 2007.

6. Teodoriu, C. Drilling Hydraulics, on behalf of NExT Training Center for PETROM S.A., Ploiesti, Romania, May 2008.

7. Teodoriu, C. Casing Design, on behalf of NExT Training Center for PETROM S.A., Ploiesti, Romania, May 2008.

8. Teodoriu, C. Directional and Horizontal drilling, on behalf of NExT Training Center for PETROM S.A., Ploiesti, Romania, May 2008.

9. Teodoriu, C. Drill string Mechanics, on behalf of NExT Training Center for PETROM S.A., Ploiesti, Romania, June 2008.

10. Teodoriu, C. Drilling Summer School, held at TU Clausthal, 2005 to 2010.

44

Appendix 2: List of Papers grouped by Subject

45

Drilling and completion Equipment, processes, special procedures (i.e. horizontal drilling)

Peer-reviewed Journals

1. Teodoriu, C. Rotary-Shouldered Connections Make-up Torque Calculation Considering the Effect of Contact Pressure on Thread Compound’s Friction Coefficient; OIL GAS European Magazine 4/2009.

Conferences

2. Paknejad, A.; Schubert, J.; Teodoriu, C.; Amani, M.

Sensitivity Analysis of Key Parameters in Foam Drilling Operations, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14 -17 October 2007.

3. Teodoriu, C.; Kinzel, H.

The Application of an Analytical model for the Controlled Make-up of Rotary Shouldered Connections in the Field, SPE 93777, SPE Asia-Pacific Conference 2005.

4. Teodoriu, C.; Reichetseder, P.; Marx, C.; Kinzel, H.

An introduction to Intelligent Make-up facilities: Analysis of Torque-Turn Recordings to make-up Rotary-Shouldered-Connections (RSC), Oil and Gas Magazine, 4/2002, ISSN 0342-5622.

5. Pilisi, N.; Teodoriu, C.

An Advisory System for Selecting Drilling Technologies and Methods with Application for Tight Gas Reservoirs, DGMK/ ÖGEW-Frühjahrstagung 2010, Fachbereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

6. Bai, M.; Fichter, C.; Teodoriu, C.; Reinicke, K. M.

Development of a Novel Testing Method to Perform Investiga-tions on Rock-Fluid Interaction under Geothermal Hot Dry Rock Conditions, DGMK/ÖGEW-Frühjahrstagung 2010, Fach-bereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

Critical equipment and components Peer-reviewed Journals

7. Reinicke, K. M.; Teodoriu, C.

Neue Entwicklungen in der Bohrtechnik, Akademie der Geowissenschaften zu Hannover Veröfentlichungen, Heft 25, 2005, ISBN 3-510-95943-4.

8. Teodoriu, C. Buttress Connection Resistance under Extreme Axial Compression Loads, Oil and Gas Magazine, Volume 31, 4/2005, ISSN 0342-5622.

Conferences

46

9. Teodoriu, C.; Schubert, J.

Redefining the OCTG Fatigue – A Theoretical Approach, 2007 Offshore Technology Conference held in Houston, Texas, U.S.A., 30 April–3 May 2007.

10. Teodoriu, C.; Reinicke, K. M.; Holzmann, J.; Klaws, M.

Testing of Tubular Goods – a Critical Review of Past and Actual Testing Procedures, DGMK Tagungsbericht 2005-1, Frühjahrstagung, am 28./29. April 2007.

11. Teodoriu, C.; Popa, A.; Klaws, M.

Optical system for leak measurement of PREMIUM threaded connections, SPC-2006,”Petroleum-Gas” University of Ploiesti, 17th- 19th May 2006.

12. Teodoriu, C.; Patil, P.

Evaluation of Surface Casing Resistance having Corrosion Damage, DGMK Tagungsbericht 2006-1, Frühjahrstagung, 28./29. April 2006, ISBN 3-936418-48-9, ISSN 1433-9013.

13. Ulmanu, V.; Badicioiu, M.; Teodoriu, C.

Laboratory for Testing Full Scale Oil Country Tubular Goods Subjected to Complex Loads, UPG Bulletin, 2006.

14. Teodoriu, C.; Popa, A.; Reinicke, K. M.; Rothfuss, M.

Implementation of an online monitoring system for the ITE Center for testing oilfield country tubular goods, SPC-2005,”Petroleum-Gas” University of Ploiesti, 11th- 13th May 2005.

15. Teodoriu, C. Contact pressure distribution on the thread flank of shouldered threaded connections using finite element analysis, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 29./30. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

16. Teodoriu, C.; Marx, C.; Reichetseder, P.

Analytical method to determine stress distribution in rotary shouldered connections, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 28./29. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

17. Teodoriu, C.; Reinicke, K. M.; Fichter, C.; Wehling, P.

Investigations on Casing-Cement Interaction with Application to Gas and CO2 Storage Wells, SPE 131336, EUROPEC/EAGE Annual Conference and Exhibition to be held in Barcelona, Spain, 14–17 June 2010.

18. Teodoriu, C.; Ugwu, I.; Schubert, J.

Estimation of Casing-Cement-Formation Interaction using a new Analytical Model, SPE131335, SPE EUROPEC/EAGE Annual Conference and Exhibition held in Barcelona, Spain, 14–17 June 2010.

47

Analysis and modeling of operational procedures, well control methods

Peer-reviewed Journals

19. Ulmanu, V.; Teodoriu, C.; Marx, C.; Pupazescu, A.

Untersuchung zur Belastbarkeit von SLIMHOLE-Bohrgestänge mit Doppelschulter, Erdöl, Erdgas, Kohle, Heft X, 125 Jg. 2009.

Conferences

20. Teodoriu, C.; Ohla, K.; Nielsen, W.

The future of environmental riserless drilling: dope free drill pipe connections using thread saver technology, OMAE-29459, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

21. Teodoriu, C.; Kinzel, H.; Schubert, J.

Evaluation of Real Time Torque-Turn Charts with Application to Intelligent Make-up Solutions, OMAE2007-29518, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

22. Teodoriu, C.; McDonald, H.; Bolfrass, C.

Friction Considerations in Rotary Shouldered Threaded Connections, OMAE2007-29583, The 26th International Conference on Offshore Mechanics and Arctic Engineering, San Diego, California, USA, June 10-15, 2007.

23. Ulmanu, V.; Teodoriu, C.

Fatigue Life prediction and Test Results of Casing Threaded Connection, Mecanica Ruperii, Buletinul ARMR, No. 17, July 2005, ISSN14538148.

Drilling and cementing fluids Health, Safety and Environment (HSE) aspects Thread compounds, fluid rheology, formation damage

Conferences 24. Ibeh, C.;

Schubert, J.; Teodoriu, C.

Methodology for Testing Drilling Fluids under Extreme HP/HT Conditions, AADE 2008 conference held April 8-9 at the Hilton (formerly Wyndham) Greenspoint Hotel in Houston, Texas, 2008.

25. Ibeh, C.; Teodoriu, C.; Schubert, J.; Gusler, W.; Harvey, F.

Investigation on the Effects of Ultra-High Pressure and Temperature on the Rheological Properties of Oil-Based Drilling Fluids, AADE 2008 conference to be held April 8-9 at the Hilton (formerly Wyndham) Greenspoint Hotel in Houston, Texas, 2008.

26. Teodoriu, C.; Schubert, J.; Vivek, G.; Ibeh, C.

Investigations to Determine the Drilling Fluid Rheology Using Constant Shear Rate Conditions, presented at the SPE IADC conference, held in Berlin, Germany, June, 2008.

48

27. Ghofrani, R.; Teodoriu, C.; Stekolschikov, K.

Swelling cement induced forces and experimental swelling pressure, DGMK Tagungsbericht 2004-2, Frühjahrstagung, 28./29. April 2004, ISBN 3-936418-17-9, ISSN 1433-9013.

Advanced drilling technology Geothermal technology

Peer-reviewed Journals

28. Teodoriu, C.; Falcone, G.

Increasing Geothermal Energy Demand: the Need for Urbanization of the Drilling Industry, Bulletin of Science, Technology & Society 2008, volume 28, No. 3, Special Issue, Part 2: The Age of Alternative Energy: Is the Future Renewable?, DOI:10.1177/0270467608315531, 16 April 2008.

29. Teodoriu, C.; Falcone, G.

Comparing completion design in hydrocarbon and geothermal wells: the need to evaluate the integrity of casing connections subject to thermal stresses, Geothermics journal, doi:10.1016/ j.geothermics.2008.11.006, 9 January 2009.

Conferences

30. Teodoriu, C.; Falcone, G.; Fichter, C.

Drilling Beyond Oil and Gas: a Discussion about How Technology will Change to Enhance Drilling in Urban Areas, DGMK/ÖGEW-Frühjahrstagung 2009, Fachbereich Aufsu-chung und Gewinnung Celle, 27./28. April 2009.

31. Fichter, C.; Reinicke, K. M.; Teodoriu, C.

Tiefbohrtechnik – Erschließung von tiefgeothermischen Wärmespeichern, GeoTHERM 2009, Offenburg, April 2009.

32. Teodoriu, C.; Fichter, C.; Falcone, G.

Drilling Deep Geothermal Reservoirs: the Future of Oil and Gas Business, Beitrag “Der Geothermiekongress 2009” Bochum, Germany, 17.-19. November 2009.

33. Fichter, C.; Reinicke, K. M.; Teodoriu, C.; Wang, Y.; Lotz, U.

Influences and limiting parameters of casing design on the success of hydrogeothermal projects - a new approach for the design of geothermal wells, Beitrag “Der Geothermiekongress 2009” Bochum, Germany, 17.-19. November 2009.

34. Reinicke, K. M.; Ostermeyer, G. P.; Overmeyer, L.; Teodoriu, C.; Thomas, R.

Der Niedersächsische Forschungsverbund Geothermie und Hochleistungsbohrtechnik (gebo), Energie und Rohstoffe 2009, Goslar, 2009.

49

35. Teodoriu, C.; Falcone, G.

Investigation of Drilling Problems in Gas Hydrate Formations, 27th International Conference on Offshore Mechanics and Arctic Engineering (OMAE 2008), Estoril, Portugal, 15-20 June 2008.

36. Teodoriu, C.; Falcone, G.

Oil and Gas expertise for Geothermal Exploitation: the Need for Technology Transfer", SPE 113852, SPE EUROPEC Conference, ROME, 2007.

37. Teodoriu, C.; Falcone, G.

Fatigue Life Prediction of Buttress Casing Connection Exposed to Large Temperature Variations, 33rd Stanford Workshop on Geothermal Reservoir Engineering, Fisher Conference Center on the Stanford University campus, 28-30 January 2008.

38. Teodoriu, C.; Falcone, G.

Comparison of Well Completions used in Oil/Gas Wells and Geothermal Wells: a New Approach to Technology Transfer, 33rd Stanford Workshop on Geothermal Reservoir Engineering, Fisher Conference Center on the Stanford University campus, 28-30 January 2008.

39. Teodoriu, C.; Fichter, C.

Anpassung der Tiefbohrtechnik aus der Erdöl-/Erdgasindustrie an die Tiefengeothermie, 3. Fachtagung, Geothermische Technologien, – Technologien zur Nutzung der tiefen Geo-thermie und ihre Integration in Energieversorgungssysteme, 23./24. März 2010.

40. Gedzius, I.; Teodoriu, C.; Fichter, C.; Reinicke, K. M.

Betrachtung von Korrosionsschäden an Casings im Primär-kreislauf geothermischer Anlagen, DGMK/ÖGEW-Frühjahrs-tagung 2010, Fachbereich Aufsuchung und Gewinnung, Celle, 12./13. April 2010.

Advanced teaching and tele-teaching

Conferences

41. Teodoriu, C.; Reinicke, K. M.; Falcone, G.

Real-Time Long Distance Teaching: An Overview of Two Years of Tele-Teaching between Texas and Germany, SPE 124748, paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

42. Teodoriu, C.; Falcone, G.

How real-time monitoring technology will change the future of laboratory classes, Teaching with Technology Conference 2007, Held at Texas A&M University, College Station, 2007.

50

43. Teodoriu, C.; Baruffet, M.; Epps, M. L.; Falcone, G.; Reinicke, K. M.

Real-Time Long Distance Teaching: The Texas A&M DL Experience, SPE, paper was accepted for presentation at the 2010 SPE Annual Technical Conference and Exhibition held in Florence, Italy, Sept. 2010.

44. Teodoriu, C.

Hands-on Teaching as Part of Petroleum Engineers Edu-cation: An International Experience, SPE, paper was accepted for presentation at the 2010 SPE Annual Technical Conference and Exhibition held in Florence, Italy, Sept. 2010.

51

Production Operations and Workover

Equipment to improve recovery efficiency, multiphase flow in pipelines

Peer-reviewed Journals

45. Falcone, G.; Teodoriu, C.; Reinicke, K. M.; Bello, O. O.

Multiphase Flow Modeling Based on Experimental Testing: A Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments, (original SPE 110116) SPE Projects, Facilities & Construction, Volume 3, Number 3, September 2008, pp.1-10.

46. Bello, O. O.; Reinicke, K. M.; Teodoriu, C.

Experimental Study on Particle Behavior in a Simulated Gas-Oil-Sand Multiphase Production and Transfer Operation. Proceedings of the 2006 ASME Joint US - European Fluids Engineering Division Summer Meeting and Exhibition, Miami, FL, USA, July 17-20 2006.

47. Surendra, M.; Falcone, G.; Teodoriu, C.

Investigation of Swirl Flows Applied to the Oil and Gas Industry, (original SPE 115938), SPE Projects, Facilities & Construction Journal, August 2008.

48. Bello, O. O.; Reinicke, K. M.; Teodoriu, C.

Particle Holdup Profiles in Horizontal Gas-Liquid-Solid Multi-phase Flow Pipeline, Chemical Engineering &Technology, Vol 28, No. 12, November 2005, ISSN 0930-7516.

Conferences

49. Solomon, F.; Falcone, G.; Teodoriu, C.

The Need to Understand the Dynamic Interaction Between Wellbore and Reservoir in Liquid Loaded Gas Wells, 27th International Conference on Offshore Mechanics and Arctic Engineering (OMAE 2008), Estoril, Portugal, 15-20 June 2008.

50. Teodoriu, C.; Falcone, G.

Low-Cost Completion for Steam Injections Wells: Theoretical and Experimental Results, SPE 110454, 2008 SPE Improved Oil Recovery Symposium to be held 21-APR-08 to 23-APR-08 in Tulsa, OK.

51. Bello, O. O.; Falcone, G.; Teodoriu, C.

Experimental Validation of Multiphase Flow Models and Testing of Multiphase Flow Meters: a Critical Review of Flow Loops Worldwide, International Multiphase Flow Conference and Exhibition 2007, Bologna, 2007.

52. Falcone, G.; Teodoriu, C.; Reinicke, K. M.; Bello, O. O.

Multiphase Flow Modeling Based on Experimental Testing: a Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments, ATCE 2007.

53. Ene, C. D.; Teodoriu, C.; Arvunescu, M.

About Kinematics of Reciprocal Pumps without Pulsation, Bul. Universitatii „Petrol-Gaze“ Ploiesti, Vol. XLVII, Nr. 6, 1998.

52

54. Popovici, A.;

Ene, C. D.; Olteanu, M.; Teodoriu, C.

Developing a Dynamical Model of a Sonic Pump Unit, Bul. Universitatii „Petrol-Gaze“ Ploiesti, Vol. XLVII, Nr. 6, 1998.

Tubular stress and strain, tubular integrity over field life, sealability of tubular compo-nents, temperature effects, corrosion issues

Peer-reviewed Journals

55. Teodoriu, C.; Ulmanu, V.; Badicioiu, M.

Casing Fatigue Life Prediction Using Local Stress Concept: Theoretical and Experimental Results, SPE 110785, 2008 SPE Western Regional Meeting held in Los Angeles, CA., 31 March to 02 April 2008.

56. Teodoriu, C.; Badicioiu, M.

Sealing Capacity of API Connections - Theoretical and Experimental Results, SPE 106849, SPE Drilling and Completion Journal, March 2009.

57. Teodoriu, C.; Dinulescu, V.

Study of Casing Loads in Steam Stimulated Wells, ICPT Campina, 1997.

58. Teodoriu, C.; Dinulescu V.

A Discussion about Casing Performance Under Thermal Cycling Conditions, The first Huf’n’Puff Project in Romania, ICPT Campina, 1997.

Liquid loading in gas wells

Peer-reviewed Journals

59. Zhang, H.; Falcone, G.; Valko, P.; Teodoriu, C.

Numerical Modeling of Fully-Transient Flow in the Near-Wellbore Region During Liquid Loading in Gas Wells, SPE-122785-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference held in Cartagena, Colombia, 31 May to 03 June 2009.

60. Park, H. Y.; Falcone, G.; Teodoriu, C.

Decision matrix for liquid loading in gas wells for cost/benefit analyses of lifting options, submitted to the Journal of Natural Gas Science and Engineering in June 2008.

Conferences

61. Chava, G.; Falcone, G.; Teodoriu, C.

Plunger Lift Modeling Towards Efficient Liquid Unloading in Gas Wells, SPE 124515, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

62. Solomon, F.; Falcone, G.; Teodoriu, C.

Critical Review of Existing Solutions to Predict and Model Liquid Loading in Gas Wells, SPE 115933, presented at the 2008 SPE Annual Technical Conference and Exhibition, Denver, Colorado, U.S.A., 21-24 September 2008.

53

Recovery of hydrocarbons Optimization of geothermal energy recovery

Peer-reviewed Journals

63. Falcone, G.; Harrison, B.; Teodoriu, C.

Can We Be More Efficient in Oil and Gas Exploitation? A Review of the Shortcomings of Recovery Factor and the Need for an Open Worldwide Production Database, Scientific Journals International, Journal of Physical and Natural Sciences, Volume 1, Issue 2, 2007.

Conferences

64. Ehsan, D.; Kegang, L.; Teodoriu, C.; McCain, W. D. Jr.; Falcone, G.

More Accurate Gas Viscosity Correlation for Use at HPHT Conditions Ensures Better Reserves Estimation, SPE 124734, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

65. Davani, E.; Kegang, L.; Teodoriu, C.; McCain, W. D.; Falcone, G.

Inaccurate Gas Viscosity at HP/HT Conditions and its Effect on Unconventional Gas Reserves Estimation, SPE-122827-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference be held in Cartagena, Colombia, 31 May to 03 June 2009.

66. Ehsan, D.; Ling, K.; Teodoriu, C.; McCain, W. D. Jr.; Falcone, G.

More Accurate Gas Viscosity Correlation for Use at HPHT Conditions Ensures Better Reserves Estimation, SPE 124734, This paper was prepared for presentation at the 2009 SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 4–7 October 2009.

67. Fernandez, J.; Falcone, G.; Teodoriu, C.

Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells, SPE-122786-MS, presented at the 2009 SPE Latin American & Caribbean Petroleum Engineering Conference to be held in Cartagena, Colombia, 31 May to 03 June 2009.

68. Teodoriu, C.; Falcone, G.; Espinel, A.

Letting Off Steam and Getting Into Hot Water – Harnessing the Geothermal Energy Potential of Heavy Oil Reservoirs, World Energy Congress 2007, ROME, 2007.

69. Haghshenas, A.; Amani, M.; Teodoriu, C.

Feasibility Study of Steam Injection in one of Iranian Naturally Fractured Heavy Oil Fields, 150 Years of the Romanian Petroleum Industry: Tradition & Challenges, Bucharest, 14 -17 October 2007.

54

Appendix 3: List of selected papers for this Habilitation (Paper full length is presented)

55

Ulmanu, V.; Teodoriu, C.; Marx, C.; Pupazescu, A.

Untersuchung zur Belastbarkeit von SLIMHOLE-Bohrgestänge mit Doppelschulter, Erdöl, Erdgas, Kohle, Heft X, 125 Jg. 2009.

Teodoriu, C. Rotary-Shouldered Connections Make-up Torque Calculation Con-sidering the Effect of Contact Pressure on Thread Compound’s Friction Coefficient; OIL GAS European Magazine 4/2009.

Teodoriu, C.; Falcone, G.

Comparing completion design in hydrocarbon and geothermal wells: the need to evaluate the integrity of casing connections subject to thermal stresses, Geothermics journal, doi:10.1016/j.geothermics. 2008.11.006, 9 January 2009.

Teodoriu, C. Buttress Connection Resistance under Extreme Axial Compression Loads, Oil and Gas Magazine, 4/2005, Volume 31, ISSN 0342-5622.

Teodoriu, C.; Badicioiu, M.

Sealing Capacity of API Connections - Theoretical and Experimental Results, SPE 106849, SPE Drilling and Completion Journal, March 2009.

Teodoriu, C.; Falcone, G.

Increasing Geothermal Energy Demand: the Need for Urbanization of the Drilling Industry, Bulletin of Science, Technology & Society 2008, volume 28, No. 3, Special Issue, Part 2: The Age of Alternative Energy: Is the Future Renewable?, DOI:10.1177/0270467608315531, 16 April 2008.

Reinicke, K. M.; Teodoriu, C.

Neue Entwicklungen in der Bohrtechnik, Akademie der Geowissen-schaften zu Hannover Veröffentlichungen, Heft 25, 2005, ISBN 3-510-95943-4.

Bello, O. O.; Reinicke, K. M.; Teodoriu, C.

Particle Holdup Profiles in Horizontal Gas-Liquid-Solid Multiphase Flow Pipeline, Chemical Engineering &Technology, Vol 28, No. 12, November 2005, ISSN 0930-7516.

Fernandez, J.; Falcone, G.; Teodoriu, C.

“Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells”, SPE Projects, Facilities & Construction Journal, June 2010.

Falcone, G.; Harrison, B.; Teodoriu C.

Can We Be More Efficient in Oil and Gas Exploitation? A Review of the Shortcomings of Recovery Factor and the Need for an Open Worldwide Production Database, Scientific Journals International, Journal of Physical and Natural Sciences, Volume 1, Issue 2, 2007.

G

Ct

CD

a

ARAA

KGWWCL

1

ooWtoGwp

hmlmdt2s

0d

ARTICLE IN PRESSG ModelEOT-763; No. of Pages 9

Geothermics xxx (2009) xxx–xxx

Contents lists available at ScienceDirect

Geothermics

journa l homepage: www.e lsev ier .com/ locate /geothermics

omparing completion design in hydrocarbon and geothermal wells: The needo evaluate the integrity of casing connections subject to thermal stresses

atalin Teodoriu, Gioia Falcone ∗

epartment of Petroleum Engineering, Texas A&M University, 3116 TAMU, College Station, TX 77843, USA

r t i c l e i n f o

rticle history:eceived 25 April 2008ccepted 14 November 2008vailable online xxx

eywords:eothermalell completion designell integrity

a b s t r a c t

Tapping for geothermal energy very often requires deep drilling in order to access high-temperatureresources. This type of drilling is expensive and is financed by the operator with a long period of debtservice before costs can be recovered from the energy sale (heat, electricity or a combination of both).Drilling costs are only a part of the total well expenditure. Tubulars can double the total well cost, espe-cially when complex well completions are required. Together, drilling and well completions can accountfor more than half of the capital cost for a geothermal power project. A comparison is made of the differ-ent completions used for oil, gas and geothermal wells, and geothermal well completion requirementsare discussed. Special attention is given to the thermal stresses induced by temperature variations inthe casing string of a geothermal well. When the induced thermal stresses exceed the yield strength of

asing fatigue

ow-cycle fatigue the casing material, the fatigue behavior of the latter can be defined as low-cycle fatigue (LCF). The con-nection threads in the casing body amplify the local stresses and lower the LCF resistance. A theoreticalapproach is presented to evaluate that parameter, and calculations are compared with preliminary resultsfrom experiments on large-diameter Buttress connections, which are commonly used in geothermal wellcompletions. It is shown that under extreme loads the LCF resistance of the Buttress thread connectioncan be as low as 10 cycles.

. Introduction

For about 100 years, deep drilling has been associated with theil and gas industry. With the advent of geothermal energy devel-pment, the drilling of deep wells has also become a requirement.hile drilling for hydrocarbons leads to a fast cash return, thanks to

he revenues from the sale of oil and gas, geothermal projects maynly break even many years after the wells have been completed.eothermal drilling costs can be “2–5 times greater than oil and gasells of comparable depths” and can “account for 42–95% of totalower plant costs” (Augustine et al., 2006).

Today’s drilling processes for the exploitation of oil and gasave been optimized to achieve affordable drilling costs (dollar pereter), though they are only part of the total well cost. The tubu-

ars (i.e. casings and tubings) used in the completion of an oil wellay double its total cost in the case of a complex well design. The

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

iameter of the last casing string in a geothermal well, often termedhe production casing, is commonly 9 − 5/8 in. (244 mm) (Teodoriu,005). Such a large diameter pipe requires a correspondingly largerurface casing and a 13 − 3/8 in. (340 mm) diameter is common-

∗ Corresponding author. Tel.: +1 979 847 8912; fax: +1 979 845 1307.E-mail address: [email protected] (G. Falcone).

375-6505/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.oi:10.1016/j.geothermics.2008.11.006

© 2008 Elsevier Ltd. All rights reserved.

place in the USA and Japan (Bohm, 2000; Jotaki, 2000; Williamsonet al., 2001).

Any well drilled deeper than 3 km would present temperaturesin line with those typical of higher enthalpy geothermal resources(Tenzer, 2001). However, this paper focuses on well integrity issuesfor hot wells, independently of their depth. In Europe, the major-ity of the geothermal wells which have been drilled to depthsgreater than 4000 m are completed with surface casing diametersof 18 − 5/8 in. (473 mm) or greater (Tenzer, 2001). The large diam-eter of production casings is a consequence of the amount of fluidsthat needs to be produced from geothermal wells. For large installedpower systems, production diameters of 13 − 3/8 in. (340 mm) arerequired, but such diameter requirements strongly affect well costs.

Over the operating life of a well, its casing string is generallysubject to external loads that can be considered as static or quasi-static. Current industry design standards, such as those issued bythe American Petroleum Institute (API), considers the casing stringto be statically loaded, yet it can be subject to variable loads dueto changes in temperature or internal pressure during geothermal

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

operations. As the movement of the casing is restricted by the pres-ence of a cement sheath, temperature variations induce thermalstresses in the casing string, which may exceed the yield strengthof the casing material. Thus, the fatigue behavior of the casing mate-rial during the operational life of a well can be classified as low-cycle

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

2 C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx

Nomenclature

aM correlation coefficient [aM = (M + 1)2 − 1]b fatigue strength exponentc fatigue ductility exponente average strainE Young’s modulus (MPa)K′ cyclic hardening coefficientK total intensity factorKts intensity factor for stressKte intensity factor for strainM factor depending on material characteristicsn′ cyclic hardening exponent2N number of cycles to failure (N—semicycles)PSWT damage parameterS average stress (MPa)S0 thermal induced stress (MPa)

Greek letters˛ expansion coefficient [m/(m ◦C)]�T differential temperature to which casing is exposed

(◦C)εa total strain amplitudeεa,e elastic strain amplitudeεa,p plastic strainε′

f fatigue ductility coefficientεl local strain�a total stress amplitude (MPa)� ′ fatigue strength coefficient

fba

ntola1nwC

2

weobgGbbga

djid

f�l local stress (MPa)�m median stress (MPa)

atigue (LCF). The presence of geometrical variations in the casingody such as the connection threads will amplify the local stresses,nd reduce the LCF resistance of the casing.

The surface casings of deep geothermal wells are exposed to sig-ificant temperature variations during drilling, which may affectheir subsequent integrity. The size of the surface casing dependsn that of the production casing and also on the drilling chal-enges presented by that particular well. The following theoreticalnd experimental work focuses on the fatigue resistance of an8 − 5/8 in. (473 mm) diameter casing with Buttress thread con-ections, which is a common size for surface casing in geothermalells, as reported by several authors (Brunetti and Mezzetti, 1970;arden et al., 1983; Chiotis and Vrellis, 1995; Tenzer, 2001).

. Geothermal resources

Fig. 1 shows the geothermal temperature gradient for selectedorld regions, with an average of 3 ◦C/100 m. As electricity gen-

ration using flash technology requires temperatures in excessf 180 ◦C; binary technology can be used for fluid temperaturesetween about 100 and 180 ◦C. The depth range for an economiceothermal project can be inferred from Fig. 1. According torimsson (2007), the temperature at depths of 4–5 km could rangeetween 200 and 300 ◦C in Europe, 300 and 400 ◦C in the USA ande greater than 500 ◦C in Japan. This paper addresses areas with aeothermal gradient in the region of 3 ◦C/100 m, corresponding tomaximum of 250 ◦C of temperature variation seen by the casing.

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

The completions of deep geothermal wells are extremelyemanding because of the high pressure imposed on the well if sub-

ected to hydraulic fracturing operations, the need for one or morentermediate casings to reach the target, the high temperature atepth and the large temperature variations along the well. In the

Fig. 1. Geothermal gradient in different regions (modified after Rogge, 2004). HDR:hot dry rock systems.

next section, the complexity of geothermal well completions willbe discussed vis-à-vis some typical oil and gas completion designs.

3. Comparison of completions used in oil/gas producersand geothermal wells

It is difficult to classify geothermal wells, as most of the exist-ing geothermal projects are custom-designed to accommodate thelocal conditions. However, a classification may be attempted on thebasis of the type of geothermal energy resources. It has been sug-gested that they should be classified by temperature (Dickson andFanelli, 1990) or in a way that reflects their ability to do thermody-namic work (Lee, 2001).

The authors have chosen the following criteria to perform acomparison of well completions used in oil/gas and in geothermalfields:

• Wellhead and surface equipment are excluded.• Tubulars, connections, and well integrity factors are accounted

for.• Three typical well completions are assumed, representative of

a geothermal producing well, a deep gas well and a heavy oilproducer.

The large diameter of production casings in geothermal wells isa consequence of the high volume and elevated enthalpy of the flu-ids being extracted. Large, high flow rate pumps (line shaft pumpsor electrical submersible) must be accommodated in geothermalwells that require external energy to extract the hot water from thereservoir. This is true for binary wells (temperature < 180 ◦C) andalso for deep Enhanced Geothermal Systems (EGS), where inducedflow circulation of an external fluid is necessary. Fig. 2 shows anexample of completions diagram for a geothermal producing wellrequiring pumping, based on the analysis of several geothermalprojects worldwide. The setting depth of the 9 − 5/8 in. (244 mm)casing is calculated so that the downhole pump is completely sub-merged at maximum flow rate. This depth may vary depending onthe characteristics of the specific geothermal reservoir. The mainchallenges for this type of completions are: the quality and long-term behavior of the cement, the selection of the appropriate casinghanger (able to withstand high temperatures) and the evaluation

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

of thermally induced loads. Casing fatigue and cement integrity arethe key issues for geothermal wells since the desired lifetime ishigher than for oil and gas. For example, Carden et al. (1983) pointedout that despite the continuous casing design improvements, therestill remain unknown factors related to casing fatigue and cement

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx 3

Conductor

Surface

Intermediate(s)

Production (9 5/8 in)

Production Liner (7 in)

ESP

Tubing

FE

id

pBueisah

uCKRVerSHLneaa1

(sgtgfltsatc

and Mezzetti, 1970; Ellis, 1979; Allen et al., 2000) of casing fail-ures for high-temperature applications, the importance of casingfatigue analysis for these wells is still not widely recognized. Theproduction of heavy oils usually requires the injection of steam,

Conductor

Surface

Tubing and downhole

ig. 2. Example of a completion diagram for a geothermal well requiring pumping.SP: electrical submersible pump.

ntegrity that could potentially cause problems in long-term pro-uction operations.

The following thread types are used in geothermal well com-letions: API Round (Brunetti and Mezzetti, 1970; Ragnars andenediktsson, 1981; Chiotis and Vrellis, 1995), virtually discontin-ed today; API Buttress (Brunetti and Mezzetti, 1970; Nicholsont al., 1982; Carden et al., 1983; Chiotis and Vrellis, 1995), whichs known for its high axial tensional strength, but low compres-ive strength; premium connections (Carden et al., 1983), whichre mostly used for production casing only, due to the associatedigh costs.

The following casing grades have been reported as commonlysed in geothermal applications: J-55 (Brunetti and Mezzetti, 1970;arden et al., 1983; Chiotis and Vrellis, 1995), usually replaced by-55 for deep applications; N-80 (Brunetti and Mezzetti, 1970;agnars and Benediktsson, 1981; Carden et al., 1983; Chiotis andrellis, 1995; Witcher, 2001), usually replaced by L-80 in the pres-nce of H2S (Lazzarotto and Sabatelli, 2005.); C-95, which hasecently been replaced by T-95, though some authors also report-95 (Carden et al., 1983) and rarely P-110, in the absence of2S. For extreme environments, 9 Chrome L-80 and 13 Chrome-80 can be utilized. Despite the often prohibitive costs, tita-ium (Beta-C Titanium) has been used for severe conditions (Pyet al., 1989). Casing pre-tensioning is reported by many authorss a common practice for geothermal well completions (Brunettind Mezzetti, 1970; Carden et al., 1983; Chiotis and Vrellis,995).

Fig. 3 shows the completion of a typical gas well, where a 7 in.178 mm) production casing is used to ensure good downhole acces-ibility and high well productivity. The first difference betweeneothermal and gas wells and is that, in the latter, the comple-ion may have the same diameter from bottomhole to surface. Theas well uses production tubing string to transport the reservoiruids to surface, so the casing string is not in direct contact with

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

he reservoir fluids, nor is it directly exposed to the reservoir pres-ure. Generally, completion fluids containing corrosion inhibitorsre placed into the annulus between tubing and casing to protecthe tubulars. It should be noted that the conventional gas wellompletion illustrated in Fig. 3 is not representative of a high-

Fig. 3. Schematic completion diagram for a typical gas well.

pressure-high-temperature (HPHT) gas well, where completion canbe 5–10 times more expensive.

A typical completion diagram for a heavy oil producer is pre-sented in Fig. 4. The main feature of these wells is the need for casingstring pre-tension, which limits the effects of axial compressiveforces due to thermal expansion. Despite many reports (Brunetti

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

Production (7 in or 5 in)

Fig. 4. Schematic completion diagram for a heavy oil producer.

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

4 C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx

Table 1Key completions parameters for geothermal, gas, and heavy oil wells.

Parameter Geothermal Gas Heavy oil

Temperature (◦C) 80–250 60–150 60–350a

Depth (m) 1000–5000 (or more) 3000-6000 300–1200Production casing ODb 9 − 5/8 in. and 7 in. liner 9 − 5/8 in. with crossover to 7 in. 5 or 7 in.Connection typec API Buttress Premium API long/Buttress

coverothermnnect

worcrpf

taftlbc

4

f

••

S

astoo(cpfsamtTs

poyc

mg

The LSSC method uses the local stress state to determine thefatigue resistance of materials. It allows the evaluation of LCFresistance of components having notches (which act as stress con-centration zones) using the results of uniaxial tests performed on

Table 2

a The high-temperature values correspond to the injected steam in thermal oil reb Larger casing outer diameters (OD’s) are reported for high power generation gec With the development of oil country tubular goods (OCTG), market premium co

hich is used to reduce the in situ oil viscosity, thus making theil more mobile. Steam injection increases the temperature of theeservoir fluids to be produced, which may represent an operationalonstraint from the well completion point of view. As heavy oileservoirs tend to be found at relatively shallow depths, the com-letions for heavy oil wells have different requirements from thoseor deep hydrocarbon wells.

Table 1 shows a selection of well completions parameters for thehree well completion types that have been discussed above. Thebove classification has shown that the common casing diametersor geothermal well are larger than 9 − 5/8 in. (244 mm), some-ime exceeding 13 − 3/8 in. (340 mm). Surface casings are generallyarger than 18 − 5/8 in. (473 mm) for such applications. It has alsoeen shown that temperature variation may exist and therefore theasing string is subjected to variable tension/compression loads.

. Thermally induced fatigue

In order to determine the influence of temperature on casingatigue, the following assumptions were made:

the casing cannot move in the cement sheath,there are no radial constraints as they would increase the averagestress, andthe induced average stresses remain in the elastic domain, butthe local stress values are in the plastic domain.

The induced thermal stresses (S0) are given by

0 = ˛E �T (1)

If the casing string is pre-tensioned, the total stress becomes anlgebraic sum of thermally induced stresses and the existing pre-tress state in the casing. Commonly, the pre-tensioning is intendedo reduce or eliminate axial compressive loads. However, a sectionf the casing must be cemented without any pre-stress state inrder to obtain the pre-tensioning state. For example, Carden et al.1983) reported that the best option for their casing design was toement the first 300 m prior to exposing the casing to a successivere-tensioning and cementing sequence. Thus, Eq. (1) is applicableor non pre-stressed sections. Additionally, fatigue becomes moreevere in symmetric alternating cycles and the effect of cementround the casing is to increase the thermal stresses. Althoughany authors have investigated this problem, it is not easy to quan-

ify these effects, especially when cement quality is not measurable.hus, it is not unreasonable to use Eq. (1) to determine the thermaltresses.

Material properties such as the yield strength change with tem-erature. This implies that, at elevated temperatures, the resistancef a casing is lower than at ambient temperature. Fig. 5 shows the

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

ield strength variation with temperature for different grades of oilountry tubular goods (OCTG) (Lari, 1997).

Table 2 shows the temperature variation that induces ther-al stresses equal to material yield strength for commonly used

rades of steel pipe. The values presented in the table were cal-

y processes.al installations.

ions tend to be extensively used for production casing in geothermal applications.

culated using Eq. (1) and the following values for steel expansioncoefficient (˛) and steel Young’s modulus (E): ˛ = 12.5 × 10−6 ◦C−1,E = 2.05 × 105 MPa. The table shows that the temperatures in anyhigher enthalpy geothermal well will cause plastic deformationeven in casing strings made of high-grade steel. It is also importantto note that, for low-enthalpy geothermal wells with temperaturesbelow 120 ◦C, a J55 steel-grade pipe is adequate. In the latter case,the well depth and the required collapse resistance are the param-eters that will dictate the appropriate selection of steel grade.

A geothermal well is subjected to cooling and heating sequencesduring its life. The thermal expansion of the casing occurs when hotfluids start flowing up the well. Repeated cooling of the wellborefor logging or other remedial activities has long been known tocause well integrity problems. Cyclic cooling and heating exposesthe casing string to cyclic tension–compression loading that maylead to casing failure. The following section presents a study thatattempts to evaluate casing fatigue using the local stress–strainconcept (LSSC), a method widely used in civil engineering and aero-nautics (Ulmanu and Ghofrani, 2001).

5. Casing fatigue: theoretical background

Fatigue represents any effect on materials due to repeated cyclesof stress. The material shows no obvious sign or warning priorto failure. It is known that cracks are produced at a later stageof fatigue, but more likely those cracks do not affect the totalsystem deformation. In some cases, like drill pipe, intensive non-destructive inspection allows the avoidance of catastrophic failuresby detecting the small cracks. Casings are normally cemented intoa well and allow no control after being set in place. Fatigue maybe classified as low-cycle fatigue and multi- or high-cycle fatigue.A typical high cycle fatigue is represented by drill pipe fatigue indeviated wells. Unlike drill pipe, the casing may be exposed bothto low-cycle as well as to high-cycle fatigue. Low-cycle fatigue is acommon type of failure when the applied loads induce high stressesin the metallic material, greater than material yield strength. Thenumber of cycles may vary from as low as 10 up to 100. High-cyclefatigue will require a large number of cycles to failure. Usually fora minimum of 106 cycles, fatigue resistance is considered to besufficient for metallic components in order to avoid catastrophicfailures. Tubular fatigue is not an unknown failure mode, but for along time it was considered as being unimportant for well casingsand tubings.

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

Temperature at which the induced stress is equal to the yield strength of the material.

Grade J55 N80 P105 P110

Temperature change, �T (in ◦C) 157 222 286 310Temperature-corrected yield strength (in MPa) 392 553 712 771

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx 5

0

100

200

300

400

500

600

700

800

900

1000

300 400 500

emperature (ºC)

Yie

ld s

tren

gth

(MP

a)J55

N80

P105

P110

s (OCTG) grades as a function of temperature (after Lari, 1997).

snctssrnt

weosctscrlaa

R

ε

w

T

ε

bTecvibdm(

�lεlE = (KS)2 (5)

The left-hand term of the equation represents a parameter that,for different loads, describes the corresponding damage. Accordingto Ulmanu and Ghofrani (2001), the damage parameter (noted as

Symmetry axisr

t

m

0 100 200

T

Fig. 5. Yield strength of different oil country tubular good

mall-scale specimens whose dimensions are specified by currentorms and standards. Casing connections are known to have stressoncentration zones due to thread geometry. Classic fatigue estima-ion requires intensive full-scale testing. The application of the localtress concept reduces the time and cost needed for a traditionaltatistical evaluation using full-scale specimens. Two input data areequired for the application of the LSSC: the experimental determi-ation of the stress–strain curve (�/ε curve) and an evaluation ofhe local stress distribution.

The stress–strain curve is measured using pure uniaxial load,ith constant deformation cycles. A constant deformation cycle

xposes the specimen to constant amplitude deformation insteadf constant amplitude stress. The results can be represented as atress–strain diagram or a Wöhler diagram, also known as stress-ycle curve (S–N curve) (Teodoriu, 2005), and is a way to representhe cyclic behavior of materials. The higher the magnitude of thetress is, smaller the number of semicycles (N) to failure. A flattenedurve (almost horizontal) at higher N values is characteristic of fer-ous materials (steels) and it is called the fatigue limit or enduranceimit. For high stress values there is no endurance limit. This islso characteristic of casing loads in which the induced stressesre higher then the material yield strength.

The stress vs. strain dependency can be written using theamberg-Osgood correlation (Ulmanu and Ghofrani, 2001):

a = εa,e + εa,p = �a

E+

(�a

K ′

)1/n′(2)

here K′, n′, and E are empirically determined parameters.The Wöhler-type diagram (Ulmanu and Ghofrani, 2001;

eodoriu, 2005) can be drawn using the following equation:

a = εa,e + εa,p = � ′f

E(2N)b + ε′

f(2N)c (3)

The experimental determination of � ′f, ε′

f, b and c is requiredecause the stress–strain curve differs from static to cyclic loading.he b and c parameters in Eq. (3) are material constants and arexperimentally evaluated, or estimated based on casing materialharacteristics determined by a tensile test. When the external loadariation is slow, the static stress–strain load may be used without

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

ntroducing large errors. For example, Fig. 6 shows a comparisonetween static and cyclic stress–strain curve for 42 CrMo4 steel. Aetailed description of how the various parameters are to be deter-ined is presented by Ulmanu and Ghofrani (2001) and Teodoriu

2005).

Fig. 6. Static vs. cyclic stress–strain diagrams for 42CrMo4/SAE 4142, hardened steel(Teodoriu, 2005).

Fatigue determination using the local stress–strain method isbased on the cyclic behavior of materials, on the relationshipbetween external loads and local stress, and on the evaluation of thestress–strain curve. The result is represented as a Wöhler diagramor stress–strain curve for a crack having a length of 0.5–1.0 mm.

The relationship between local and average stresses and strainsis given by the Neubert equation (Ulmanu and Ghofrani, 2001):

K2 = KtsKte = �l

S

εl

e(4)

Using the notation e = S/E, Eq. (4) can be rewritten as

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

D

Fig. 7. Typical U-type notch that mimics a threaded connection (based on DIN 471,1990). D: external diameter (mm); r: notch root fillet radius (mm); m: notch width(mm); t: notch depth (mm).

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

6 C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx

ent m

P

P

u

P

saa

mt

S

bcVsa

K

c

6

tV2c

compression. This variation can be explained by the more aggres-sive bending on the thread turn for the tensile load and also bythe asymmetric geometry of the Buttress thread that has differ-ent flank angles. For Buttress threads, the flank angles are 3◦ for

Fig. 8. Geometry and mesh of a two-dimensional finite elem

SWT) is given by

SWT =√

�lεlE (6)

The so-called modified Wöhler curve of PSWT can be determinedsing the following equation:

SWT =√

� ′f2(2N)2b + � ′

fε′fE(2N)b+c (7)

By solving Eqs. (3) and (5), it is possible to determine the localtress and strain for a given load. Given �l and εl, parameter PSWTnd the number of semicycles, N, can be calculated using Eqs. (6)nd (7).

The influence of the average tension on the damage processust also be considered. The following equation can be used for

his purpose:

= aM(�m + �a) (8)

For specific notch geometries, the stress concentration factor cane calculated using empirical formulae. The equivalent notch of aasing connection thread can be modeled by using the so-called- or U-type notch. For the U-type notch (see Fig. 7), the Germantandard DIN 471 (1990) provides the following formula (assumingbending stress state):

= 1.14 + 1.08

√t

r= 1.14 + 1.08

√1.570.2

= 4.17 (9)

For standard threaded connections (nut and bolt) the stress con-entration factor is considered to be between 4 and 10 (Buch, 1988).

. Casing fatigue: numerical investigations

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

The finite element method (FEM) was used to better understandhe stress state in the threaded connection of a casing. The ANSYSersion 11.0 (University License) software package (Lethbridge,008) was used to investigate the stress induced in the threadedonnection during axial tension and compression.

ethod model of an 18 − 5/8 in. diameter casing connection.

The first goal was to estimate the stress concentration fac-tors within the complete and incomplete thread turns. Changesin material properties due to high temperature were not initiallyconsidered, as it was assumed that these factors are a function ofgeometry only.

A two-dimensional model of a N80 grade, 18 − 5/8 in. (473 mm)casing with 11.05 mm wall thickness was built (see Fig. 8), whichprovided fast and reliable results. The connection is considered tobe axially restrained at the ends, with free radial displacement. Oneend has zero degrees of freedom, while the other end has onlythe radial displacement blocked. This ensures a pure axial loaddue to the applied force. The different way in which the connec-tion responds to the loads was evaluated using a three-dimensionalmodel, as presented in Fig. 9.

The stress concentration factor, K, is defined as the maximumstress that occurs in the connection (commonly at the thread root)vs. the average stress in the casing body. Based on the FEM analysis,it was found that, for a Buttress connection K differs from tension to

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

Fig. 9. Qualitative representation of the deformation of a compression loaded (a)and tension loaded (b) threaded connection compared to the non-deformed status(c).

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx 7

Fig. 10. Stress distribution at the thread root for the compression (top) and tension (bottom) load cases, as simulated using the finite element method model.

0.00

50.00

100.00

150.00

200.00

250.00

1000100101

mber

Tem

pera

ture

var

iati

on (

ºC)

12

Measured

K = 3

K = 5

K = 4.17

K = Total intensity factor

18 − 5

tsuasb

ff

7

Ltje

f

so that the average stress inside the specimen body was equal tothe material’s minimum yield strength (550 MPa). A more detaileddescription of the testing procedure is presented by Teodoriu(2005).

Table 3Parameters used for casing fatigue calculations.

Parameter Notation Value

Nu

Fig. 11. Temperature vs. number of cycles for an

he loaded (or active) flank and 10◦ for the unloaded flank. Fig. 10hows the loaded and unloaded flanks’ positions, and thread turnsnder compression and tension loads, with high resulting stressest the thread root. In reality, the incomplete thread turns presentharp edges and therefore higher stress concentration factors maye expected.

The deformation of the thread turns under a tension load is dif-erent from that under a compression load; we obtained: K = 3.51or tension and K = 2.73 for compression.

. Discussion

The modified stress-cycle curve (S–N curve) obtained from theSSC theoretical model is given in Fig. 11. The y-axis represents

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

he maximum differential temperature to which the casing is sub-ected. Table 3 presents the parameters used for the casing fatiguestimation, which appear in Eq. (7).

Fig. 11 shows that, for extreme temperature variations, theatigue resistance of the connections is as low as 10 cycles. It

of cycles

/8 in. (473 mm) Buttress connection, grade N80.

took 12 cycles to reach fracture with the full-scale experimentaltests, corresponding to a stress concentration factor of 4.17. Theexperimental tests were conducted with a grade N − 80, 18 − 5/8 in.(473 mm) casing. The specimens were axially loaded in a testingframe, and exposed to alternating cycles of tension and compres-sion until fracture occurred. The applied axial load was calculated

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

Constant b −0.079Constant c −0.869Young’s modulus E 182 GPaMaximum strain εf 1.78Tension strength �f 720 MPa

ARTICLE IN PRESSG ModelGEOT-763; No. of Pages 9

8 C. Teodoriu, G. Falcone / Geothermics xxx (2009) xxx–xxx

m) B

capipsfcctt5pe

tmmop

ptTigttvacb

8

tft

r1

Fig. 12. Full-scale testing of an 18 − 5/8 in. (473 m

Several 18 − 5/8 in. (473 mm) specimens representing a Buttressonnection of Grade N80 casing were tested under alternatingxial tension and compression loads with slow variation (one cycleer hour). The applied loads were ±9300 kN, corresponding to an

nduced axial load equal to the yield strength of the pipe body. Alot of the applied cycles is shown in Fig. 12, which also shows thepecimen mounted in a full-scale testing frame. The first specimenailed (by rupture) at the last engaged thread of the pipe after 10ycles. The wall thickness in the failure area varied across the pipeircumference. A cross-section of the failed specimen showed thathe pipe thickness in the threaded area was only 91.5% of that ofhe rest of the pipe, leading to a nominal stress in the failure area of81 MPa, which was approximately 9% greater than the stress in theipe body. No fractures or cracks appeared in two other specimensxposed to about six cycles each.

The stress concentration factor varies from one type of connec-ion to another, as it is a function of connection geometry and its

anufacturing process. The fracture obtained during the experi-ental investigations on Buttress thread was located at the zone

f imperfect thread turns, suggesting that more attention must beaid during the thread manufacturing process.

For a high-enthalpy geothermal wells with fluid production tem-eratures between 100◦ and 250 ◦C the fatigue resistance of theested N80 Buttress connections varies between 10 and 110 cycles.his information should be considered when evaluating the min-mum project lifetime and optimizing well operations. The threadeometry, especially the incomplete thread turns, strongly affecthe value of K. For example, a lower K will increase the lifetime ofhe casing over 1000 cycles, as shown in Fig. 11. Large temperatureariations should be avoided by limiting well cooling. Maximumllowable temperature difference can be calculated using fatigueriteria as presented here or plastic deformation criteria discussedy Chiotis and Vrellis (1995).

. Conclusions and recommendations

Drilling and well completions can account for more than half ofhe capital cost for a geothermal power project. The lessons learnt

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

rom oil and gas wells may be helpful in geothermal wells comple-ions, both from the technical and the economic point of views.

The paper presented the theoretical and experimental work car-ied out to evaluate the low-cycle fatigue (LCF) resistance of an8 − 5/8 in. (473 mm) diameter casing with Buttress thread connec-

uttress connection, grade N80 (Teodoriu, 2005).

tions (typically used in geothermal well completions). The resultsshowed that, under extreme loads, the LCF resistance of this type ofconnections could be as low as 10 cycles. More full-scale exper-imental work is required to extend the validity of the resultsobtained by this study to other types of threaded connections. Also,as considerable geothermal field experience (>2000 wells) withButtress threads is available, the results presented here need to beevaluated in light of field experience.

We consider that temperature variations in geothermal wellsduring drilling, testing, repair and other operations should be keptto a minimum. For example, as it can be deduced from Fig. 11, atemperature variation of less than 80 ◦C is required to ensure thatthe total number of cycles to failure remains higher than 100. Thiswould minimize the risk of casing failure for the average life of ageothermal well.

Significant care should be taken when cyclic water injec-tion/production (huff ‘n’ puff) operations are performed, like theones discussed by Wessling et al. (in press). In addition, whenconsidering reservoir stimulation procedures one should try tominimize the injection of cold water into the (hot) geothermalwells.

References

Allen, D.M., Ghomshei, M.M., Sadler-Brown, T.L., Dakin, A., Holtz, D., 2000. Thecurrent status of geothermal exploration and development in Canada. In: Pro-ceedings of the 2000 World Geothermal Congress, Kyushu-Tohoku, Japan, pp.55–58.

Augustine, C., Tester, J.W., Anderson, B., Petty, S., Livesay, B., 2006. A comparisonof geothermal with oil and gas well drilling costs. In: Proceedings of the 31stWorkshop on Geothermal Reservoir Engineering, Stanford University, Stanford,CA, USA, pp. 5–19.

Bohm, B., 2000. Drilling geothermal well ISO. Geo-Heat Center Quarterly Bulletin 21(4), 12–19.

Brunetti, V., Mezzetti, E., 1970. On some troubles most frequently occurring ingeothermal drilling. Geothermics Special Issue 2 (Part I), 751–756.

Buch, A., 1988. Fatigue Strength Calculation. TransTech Sa, Aedermannsdorf, Switzer-land, 256 pp.

Carden, R.S., Shursen, G., Nicholson, R.W., Pettitt, R.A., Rowley, J.C., 1983. Uniqueaspects of drilling and completing hot dry rock geothermal wells. In: Proceedingsof the IADC/SPE 1983 Drilling Conference, New Orleans, LA, USA, 20–23 February,pp. 123–140.

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

Chiotis, E., Vrellis, G., 1995. Analysis of casing failures of deep geothermal wells inGreece. Geothermics 24, 695–705.

Dickson, M.H., Fanelli, M. (Eds.), 1990. Small Geothermal Resources. UNITAR/UNDPCentre for Small Energy Resources. Rome, Italy, pp. 1–29.

DIN 471, 1990. German Standard DIN 471, Sicherungsringe (Halteringe) für Wellen.Regelausführung und schwere Ausführung, 29 pp.

ING

eother

E

G

J

L

L

L

L

N

P

press. Pressure analysis of the hydromechanical fracture behaviour in stimulated

ARTICLEG ModelEOT-763; No. of Pages 9

C. Teodoriu, G. Falcone / G

llis, P.F., 1979. Failure Analysis Report: Well Production Casing Brady No. 5 Well.Geothermal Food Processors, Inc., Fernley, Nevada. Radian Corporation reportDCN 79-212-003-04 prepared for the U.S. Department of Energy, Chicago Oper-ations and Regional Office, Argonne, IL, USA, 17 pp.

rimsson, O.R., 2007. A Clean Energy Future for the United States: the Case ofGeothermal Power. Testimony by the President of Iceland Olafur Ragnar Grims-son at a Senate Hearing Committee on Energy and Natural Resources, US Senate,September, 31 pp.

otaki, H., 2000. High angle directional drilling at Takigami geothermal field. In:Proceedings of the 2000 World Geothermal Congress, Kyushu-Tohoku, Japan,pp. 2347–2352.

ari, I., 1997. Rezistenta Coloanelor de Bulrane Tubate in Sondele Supuse Procesuluide Injectioe de Abur. ICPT internal Report, Campina, Romania, 55 pp.

azzarotto, A., Sabatelli, F., 2005. Technological developments in deep drilling in theLarderello Area. In: Proceedings of 2005 World Geothermal Congress, Antalya,Turkey, p. 6.

ee, K.C., 2001. Classification of geothermal resources by exergy. Geothermics 30,431–442.

Please cite this article in press as: Teodoriu, C., Falcone, G., Comparing coevaluate the integrity of casing connections subject to thermal stresses. Ge

ethbridge, P., 2008. An overview of ANSYS academic products at release 11.0.http://www.ansys.com/academic/downloads/academic article pdf 1.pdf.

icholson, R.W., Pettit, R.A., Sims, J., 1982. Production casing for hot dry rock wellsEE-2 and EE-3. Geothermal Resources Council, Transactions 6, 219–222.

ye, D.S., Holligan, D., Cron, C.J., Love, W.W., 1989. The use of beta-C titanium fordownhole production casing in geothermal wells. Geothermics 18, 259–267.

PRESSmics xxx (2009) xxx–xxx 9

Ragnars, K., Benediktsson, S., 1981. Drilling of a 2000 m (6562 ft) borehole forgeothermal steam in Iceland. In: Proceedings of the International GeothermalDrilling and Completions Technology Conference, Albuquerque, NM, USA, 21January, pp. 4.1–14.1.

Rogge, S., 2004. Geothermische Stromerzeugung in Deutschland Ökonomie, Ökolo-gie und Potenziale. Ph.D. Thesis. Technichen Universität Berlin, Germany, 119pp.

Tenzer, H., 2001. Development of hot dry rock technology. Geo-Heat Center QuarterlyBulletin 22 (4), 14–22.

Teodoriu, C., 2005. Cercetari privind comportarea materialelor sa imbinarilor file-tate in sondele supuse procesului de injectie de abur. Ph.D. Thesis. Universitatea“Petrol-Gaze”, Ploiesti, Romania, 155 pp.

Ulmanu, V., Ghofrani, R., 2001. Verfahren zur Lebensdauerabschätzung der Tief-pumpgestänge nach dem Ortlichen Konzept (Fatigue life prediction methodfor sucker rods based on local concept). Erdöl, Erdgas, Kohle 117 (4),189–195.

Wessling, S., Junker, R., Rutqvist, Silin, D., Sulzbacher, H., Tischner, T., Tsang, C.F., in

mpletion design in hydrocarbon and geothermal wells: The need toothermics (2009), doi:10.1016/j.geothermics.2008.11.006

tight sedimentary geothermal reservoirs. Geothermics.Williamson, K.H., Gunderson, R.P., Hamblin, G.M., Gallup, D.L., Kitz, K., 2001. Geother-

mal power technology. Proceedings of the IEEE 89, 1783–1795.Witcher, J.C., 2001. Geothermal direct-use well for commercial greenhouses radium

springs, New Mexico. Geo-Heat Center Quarterly Bulletin 22 (4), 1–7.

Copyright 2007, Society of Petroleum Engineers This paper was prepared for presentation at the 2007 SPE Production and Operations Symposium held in Oklahoma City, Oklahoma, U.S.A., 31 March–3 April 2007. This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.

Abstract Hydraulic fracturing has become a state of the art stimulation technique. It has been proved over the years that significant production increase can be obtained by applying the right fracturing technique. Nowadays, the most advanced techniques of geothermal energy recovery systems widely use hydraulic fracturing. The following paper presents the experimental results of the tests carried out on four different compounds using the improved “grooved plate” method. The tests have showed a large variation of the tested thread compounds sealing capacity. Starting from the experimental results and the theoretical analysis of the API connection a useful chart was built to determine the real connection resistance, based on its initial makeup torque. The chart offers to engineers involved in the design of a fracturing process the possibility to estimate the maximum pressure that may lead to a connection leak.

Introduction Most of the published data show that a long fracture is the key to well optimum stimulation. The desired length of the fracture can be achieved using equipment capable to deliver the right pressure and fluid volume. Since the hydraulic fracturing technique can be also applied to old wells, equipped with standard API connections, the high pressures that are achieved during the pumping phase require the understanding of leak resistance of API connections. It has been also proven that during the injection phase the high pump rate may lead to additional pressure increase into the well tubulars. The time and pressure values are two key parameters that may affect the sealing capacity of the API connection. Testing the sealing capacity of a casing connection is not an easy task since it depends on many factors like: thread type and form, thread compound, ageing of the thread compound, make-up induced stresses, etc. Actually, there are no standards to evaluate the seal capacity of a thread compound. To date, three approaches have been found in the literature:

• the fixture designed during the project PRAC 88-51 that consists of two circular steel plates having spiral grove from the center to the exterior [1];

• full scale testing of threaded assembly using high load press [2] in which not only the thread compound but the entire sealing capacity of the assembly is tested;

• small size connections as described in paper [3] There are many pros and cons for each one of the methods, but testing thread compounds separately require getting off all inconsistent parameters that may affect the evaluation process. The main parameters that may affect the thread compound evaluation are the stress-strain state induced due to make-up and thread tolerances. The fixture proposed by the project PRAC 88-51 offers the advantage of comparing the threaded compounds only, by neglecting the make-up and tolerances induced errors. This is why it has been considered the use of the same experimental setup as the one described in paper [1]. The experimental setup will be presented in detail later in this paper.

Thread Compounds for Oil Country Tubular Goods (OCTG) Typical threaded compounds for OCTG are formed using base grease in which solid particles are dispersed. The grease is standard lubricating grease made of mineral oil having a metal soap as thickener (i.e. aluminum stearate). In very low amount, additives are added to the compound to improve the following properties: high pressure resistance, wear protection, corrosion protection, etc. The role of solid particles is to provide anti-galling resistance and sealing properties of the compound. Powdered metals and non-metallic particles like graphite or ceramic spheres are used as solid ingredients. Typical metals used for threaded compounds manufacturing are: lead, copper, zinc. The common non-metallic solids used for compounds are graphite, PTFE, ceramics. The so called “green dope” or environmental friendly compounds have a totally metal-free composition. Figure 1 shows a classification scheme of thread compounds after [3]. Table 1 shows the composition of some common threaded compounds used in the oil industry, including the tested thread compounds described in this paper. According to [4] the performance general requirements of threaded compounds include: consistent frictional properties, adequate lubrication properties, adequate sealing properties, physical and chemical stability both in service and in storage conditions, and properties that allow the efficient application of the compound on the connection surfaces. In addition, for

SPE 106849

Sealing Capacity of API Connections - Theoretical and Experimental Results Marius Badicioiu/UPG Ploiesti, Catalin Teodoriu/Texas A&M

2 SPE 106849

RSC threaded compounds they should lubricate the connection during the make-up runs to achieve bearing stresses (buck-up force). The sealing capacity or, according to some authors, leak tightness, is provided by the high viscosity of the threaded compound and the small free path inside of the threaded connection.

The API threads The API round thread is one of the very first standardized thread type used for casing and tubing. Being cheap, simple and easy to manufacture the API round threads have been extensively used for “low cost” wells. A proof of their importance for the oil industry is given by the extensively tests carried out by a technical advisory committee in the 90s having Mr. Phil Pattillo as Chairman [5, 6, 7]. The tests were focused on the better understanding of the mechanical behavior of the API 8 Round casing connection when subjected to service loads of assembly interference, tension and internal pressure [8, 9, 10] The API round thread is an open type thread, which means that if no other material is applied on thread prior its make-up no seal is provided. Between threads a small gap remains that must be filled in order to provide a leak tight connection. This space is filled by the thread compound that is applied on pin and box before make-up. The shape and dimensions of the API 8rd are given in Figure 2. As it can be seen in the Figure 2 the leak path consists of two spiral spaces comprising the gaps between pin thread crest and box thread root and pin thread root and box thread crest, respectively. The Buttress thread has been introduced later in order to offer a connection with a very good tension resistance. It was patented Nov. 27, 1956 by Mr. Samuel Webb and assigned to United States Steel Corporation. As stated in the patent the high leak resistance should not be expected unless the stab flank is closed. The shape of the Buttress has a different leak path than API round, usually much higher. Early API tests [8] showed no difficulties with the leak resistance, but it must be noted that at that time API compounds were used for tests only. The leak path size depends on the thread manufacturing tolerances. In order to calculate the gap volume, the minimum and maximum tolerances have been considered as presented in Annexes A and B for API round, respectively, API Buttress. The experimental setup As it has been explained before the grooved plate setup has been chosen for thread compound analysis because it allows testing of thread compounds independent from thread tolerances and stress-strain state. Later, full scale specimens with controlled geometry have been used for reference. The test was performed according to the following procedure: the grooved plate was completely filled with the dope to be tested, and then the grooved plate was assembled over the sealing plate into the hydraulic press. The center of the grooved plate was connected to the high pressure pump. Mineral hydraulic oil has been used as pressurizing medium. The pressure was slowly increased and the moment at which the dope was expelled was recorded. To build the groove plates, the groove size has been calculated

according to the real dimensions of a 5 ½” API short round thread, with a wall thickness of 7.72 mm with J55 grade. The calculated dimensions for the grooved plate are shown in Table 2. The grooved and seal plates are shown in Figure 3 and 4. Finite element analysis of the threaded connection and experimental setup In order to estimate the contact stress to be simulated between grooved plate and seal plate a finite element analysis was carried out, using ANSYS University program. The contact pressure between the thread turns plays an important role for the sealing capacity of the connections because as long as the contact pressure is higher then the pressure to be sealed the only leak path remains the spiral path between thread turns crest and roots. Same conditions must be achieved between the two plates of the experimental setup. Firstly, the 5 ½” connection has been investigated by determining the thread turn contact pressure after make-up and under make-up and axial load case. The results were similar with those reported by Asbill and Pattillo in [5, 6, 7]. According to the finite element analysis the contact pressure after the connection make-up with recommended torque is between 30 to 45 MPa and depends on how the averaging is performed. Due to local contact problems within the incomplete thread turns zone, some spots with high contact pressure have been found, see Figure 5. These values will not be considered for the further analysis. It has been also observed that the pressure on the thread flank is not uniform, as reported by Asbill and Pattillo [6], but for the experimental setup an average value has been considered. Same analysis was carried out on several Buttress connection sizes. Figure 6 shows the flank contact pressure for an 18 5/8” connection. The average contact pressure was 65 MPa. The third finite element analysis was performed on the grooved plate model in order to investigate the groove deformation due to axial load applied on the plate. The results have showed that the pressure distribution on the contact area is uniform, excepting the contact zones at the end of the groove walls (see Figure 7). Also, it has been found that the shape of the groove is changing, as presented in Figure 8. The total amount of area shrinkage for the contact pressures produced in a real connection is low. When the contact pressure increases the groove deformation becomes important, especially for the API round plate model. For a contact pressure of 50 MPa the grove area changes for the Buttress plate are of 0.85% and for the API plate of 7.3%, which represents a 9 time area change compared to Buttress. These changes justify the results and show that the API plate leak resistance increases almost linearly compared to Buttress plate leak resistance (see Figure 9). Experimental results The first experiments were focused on testing four thread compounds using the grooved plate that mimics the Buttress thread. The tested compounds are presented in Table 3. The first two thread compounds are proprietary types, therefore they will be called T1 and T2, the third one is an API Modified and the last one is an API type compound with polymers to increase the viscosity. The API Modified was

SPE 106849 3

used as reference for all tests. Figure 8 shows the leak pressure as a function of contact pressure between plates. The leak pressure is the pressure at which the thread compound is expelled from the free end of the groove. It can be seen that all tested compounds show a higher leak pressure than the API Modified. Also, it has been observed that at high contact pressures the leak pressure tends to behave asymptotic. The thread compounds T1 and T2 as well as the API with polymers show very little differences at high contact pressures between plates. This is explained by the slight deformation of the groove due to confine axial force applied to keep the plates together and the viscosity of the compounds. The second set of tests was carried out on a grooved plate that mimics the API short round thread. For these tests only the thread compound T1 has been evaluated in order to compare the results with the full scale specimens that have been doped with this type of compound. Figure 9 shows a comparison between the results obtained using Buttress type groove and the API round type groove. One explanation for the asymptotic behavior of the API Buttress leak resistance curve is the groove deformation due to axial load of the plate. As presented before, the groove deformation can reduce its volume up to 7% for API style groove and 0.8% for Buttress style groove. Since the API groove size becomes smaller and smaller with the increasing contact pressure, it is obviously why the API plate leak resistance is a direct function of contact pressure, and Buttress plate not. Practical application of the method It is a common practice to record the applied torque while running casing. Knowing the minimum value for the threaded connections applied torque, it will be easier to estimate the maximum pressure that may be applied without loosing the connection tightness. In many fracturing operations the part of the casing string that is subjected to internal pressure will have a low or zero axial tension. Therefore, the chart presented in Figure 10 has been constructed using the contact pressure calculated for the make-up torque and internal pressure case. Comparing the contact pressure inside the threaded connection to the minimum contact pressure at which the thread compound has been expulsed form the small scale setup it is possible to identify the actual leak resistance of the connection. The method does not consider the effect of tolerances and thread compound ageing. Conclusions The tests performed on four different types of compounds have showed that the API compound has the lowest leak resistance in conjunction with the API thread type. The Buttress leak resistance has an asymptotic behavior. At contact pressures higher than 100 MPa, the leak resistance is constant. The difference between API round and Buttress leak resistance consists in the contact pressure dependency of the API round leak resistance. It is recommended for API round connections to be made-up with optimum make-up torque or higher.

Acknowledgement The authors will like to thank Professor Vlad Ulmanu for his standing support and valuable advices. References 1. ***, Thread compound test being developed, Oil & Gas Journal, Spet 10, 1990 2. ISO 13678, 3. Hoenig, S. Oberndorfer M. Tightness Testing of Environmentally Friendly Thread Compounds, SPE 100220, SPE Europe/EAGE Conference, Vienna, Austria, 12-15 June 2006 4. ***, ISO13678:2000(E) 5. Report on API PRAC Project 88-51, “Investigation of Pipe Thread Compound”, June 1989 6. Report on API PRAC Project 86-51, “Investigation of Pipe Thread Compound”, May 1987 7. Report on API PRAC Project 84-51, “Investigation of Pipe Thread Compound”, December 1985 8. Report on API PRAC Project 86-53, “Investigation of Leak Resistance of API Buttress Connector”, January 1987 Table 1. Tested thread compounds composition

Product Specification

API Mod. Compound

T1 T2

Grease Base 36.0% N/A ~46% Powdered Graphite 18.0% N/A 20% Lead Powder 30.5% 0% 0% Zinc Dust 12.2% N/A 20% Copper Flake 3.3% 0% 0% Thickener Al Stearate Ca Stearate Ca StearateFluid Type Petroleum Oil Oil Density 1700 kg/m³ N/A N/A Colour Black/Brown Black Black

Table 2. Size and dimensions of the grooved plate for two different connection types

Area, [mm2] D, [mm]

Thread turn dimensions, [mm]

Tota

l th

read

cl

eara

nce

Gro

ove

area

h b ps

API round 0.049 0.056 120 0.08 0.70 1.75

API Buttress 0.312 0.310 156 0.20 1.55 3.1

Table 3. Leak resistance of Buttress type plate

Contact pressure [MPa) Thread compound 43.1 53.9 64.7 75.5 86.2 97 T1 33.2 38.8 42.0 43.8 45.0 45.2 T2 32.0 37.0 40.3 42.2 43.7 44.0 API 5A2 19.0 24.3 28.5 32.0 35.0 37.2 API with polymers 29.0 33.5 37.0 39.8 42.5 44.0

4 SPE 106849

Fig. 1. Classification Scheme of Thread compounds, after [3]

D

25

ps

b h

12,5

Ø50

- b -

Ø6

Ø200

M6

40

40

A A

A–A

- a -

Fig. 3. Shape and dimensions of the grooved and seal plates used for the experiments

p = 3,175

60°

p/2

30° 30°

h s=1

,809

75S

rs=0

,431

8s c

s=0,

508

H=2

,749

55

90°

Srn

=0,4

318

h n=1

,809

75s c

n=0,

508

H=2

.749

55

A1 = 0,049 mm2

A2 = 0,049 mm2

Fig. 2. API Round Thread Dimensions

Fig. 4. Pictures of the grooved (top) and seal (bottom) plates

More special use

Mor

e en

viro

nmen

tal f

riend

ly

SPE 106849 5

Fig. 5. Flank contact pressure of a 51/2” API round threaded connection after optimum make-up torque

Fig. 8. Leak pressure curves for API Buttress type plate

Fig. 6. Flank contact pressure of a 18 5/8” API Buttress threaded connection after optimum make-up torque and axial

tension Fig. 9. Comparison of API round and Buttress leak resistance

test results for thread compound T1

-5.00E-02

0.00E+00

5.00E-02

1.00E-01

1.50E-01

2.00E-01

2.50E-01

Gro

ove

heig

ht [m

m]

-5.00E-02

0.00E+00

5.00E-02

1.00E-01

1.50E-01

2.00E-01

2.50E-01

Gro

ove

heig

ht [m

m]

Fig. 7. Groove deformation at high contact pressure for API round type plate (left) and Buttress type plate (right)

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

5 0

4 0 5 0 6 0 7 0 8 0 9 0 1 0 0

C o n ta c t p r e s s u r e , [M P a ]

Leak

pre

ssur

e [M

Pa]

T 1T 2A P I w i th p o ly m e rsA P I 5 A 2 0

10

20

30

40

50

60

70

0 20 40 60 80 100 120 140

Contact pressure [MPa]

Leak

Pre

ssur

e [M

Pa]

API round

API Buttress

6 SPE 106849

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

0 5 10 15 20 25 30 35 40 45

Internal applied pressure, [MPa]

Con

tact

pre

ssur

e in

side

of t

he c

onne

ctio

n , [

MP

a]

Experimentally determined leak resistance

Min. test pressure according to API = 20.7 [MPa]

Mt minim = 2400 Nm

Mt optim = 3150 Nm

Mt maxim = 3950 Nm

Actual make-up torque

Estimated connection

leak resistance

Fig. 10. Chart for API connection leak resistance estimation knowing the thread compound leak resistance

Appendix A Leak path dimensions as a function of thread tolerances for API round thread

Leak path area, [mm2] Case Thread Height, [mm] A1 A2 Shape and dimensions of the thread gaps

1 hs = 1.80975 hn = 1.80975

0.0490 0.0490

2 hs = 1.80975 – 0.102 hn = 1.80975 + 0.051

0.1555 0.1555

3 hs = 1.80975 + 0.051 hn = 1.80975

0.0153 0.0153

4 hs = 1.80975 – 0.102 hn = 1.80975 – 0.102

0.0597 0.0597

8 SPE 106849

Appendix B Leak path dimensions as a function of thread tolerances for API Buttress thread

Case Tolerances Leak path area, [mm2] Shape and dimensions of the thread gaps

1 No tolerances 0.0472

2 Deviation from thread turn length and fillet radius

A1: 0.3116 A2: 0.0036 A3: 0.0036

3 Deviaiton from thread height 0.1646

Increasing Geothermal Energy Demand: TheNeed for Urbanization of the Drilling Industry

Catalin TeodoriuGioia FalconeTexas A&M University

Bulletin of Science, Technology & Society Vol. XX, No. X, Month XXXX, xx-xxDOI: 10.1177/0270467608315531Copyright © 2008 Sage Publications

Drilling wells in urban spaces requires specialtypes of rigs that do not conflict with the surroundingenvironment. For this, a mutation of the currentdrilling equipment is necessary into what can bedefined as an “urbanized drilling rig.” Noise reduc-tion, small footprint, and “good looking” rigs allhelp persuade the general public to accept the pres-ence of drilling rigs in their neighborhood. This arti-cle reviews international projects that aim tointegrate drilling with the urban infrastructures witha special focus on geothermal projects. Case studiesare presented where tailored drilling rigs and newtechnology have already been implemented. Thereview aids the analysis of the main urban-relatedtechnical aspects of modern drilling rigs. Finally, thenew trends in integrating architecture, urbanism, anddrilling rig design are discussed.

Keywords: geothermal; energy; drilling; rigs; urban-ization; environment.

Introduction to Well Drilling:Past and Present

History of Well Drilling

Drilling equipment is needed to penetrate under-ground reservoirs for water, oil, and natural gas, orinto subsurface mineral deposits. Drilling systems arealso required to exploit geothermal resources and forcivil and mining engineering applications.

According to the American Ground Water Trust(2007), there are more than 15 million homes in Americawith their own water well. More than half of the nation’s

drinking water is from municipal and private wells.Some estimates suggest that 6,000 new wells are com-pleted each week in the United States.

In 1848, the first modern oil well was drilled in Asia,on the Aspheron Peninsula northeast of Baku, by aRussian engineer named F.N. Semyenov (Talwan,Belopolsky, & Berry, 1998). In 1854, the first oil wellswere drilled in Europe, at Bóbrka, Poland by IgnacyLukasiewicz, who also designed the kerosene lamp(Arabas, 2005). In 1859, the most important U.S. oilwell was drilled in northwestern Pennsylvania byColonel Edwin Drake (Baker, 1996). This was one ofthe first successful oil wells that were drilled for the solepurpose of finding oil. Known as the Drake Well, afterthe man responsible for the well, it began an interna-tional search for petroleum, and in many ways eventu-ally changed the way we live. The Spindle Top well,drilled by Anthony Lucas in Texas in 1901, was the firstwell to be drilled by rotary drilling, to overcome geo-logical problems.

Geothermal resources for the production of electricityhave been used since the beginning of the 20th century.The first successful experiment to produce energy fromgeothermal sources was carried out in 1904 by PrincePiero Ginori Conti at the Larderello dry steam field inItaly (Lungonelli & Migliorini, 2003). Today, Larderelloprovides 10% of the world’s supply of geothermal elec-tricity. In the United States, the first large-scale geother-mal electricity-generating plant began operation in 1960,at The Geysers in Sonoma County, California (U.S.Department of Energy, 2006). 69 facilities are in opera-tion at 18 resource sites around the country.

The depth that the well has to reach is one of themain criteria that dictate the size and the specifica-tions required for a drilling rig. In the case of water

wells, for instance, the drilling depth will depend onwhere the water table is located. Also, drilling timeand depth are usually directly related: The deeper thewell, the longer it will take to drill it (although thetype of formation to be drilled—hard or soft—willalso have an impact on the overall drilling time).

Drilling Rigs

A drilling rig’s sole purpose is to get a hole from asurface location to a subsurface target at a specifieddepth in a safe and controlled manner. All the equip-ment that makes up a drilling rig is designed andmanufactured to this aim. Generally, a drilling rigconsists of five main subsystems: pipe rotation,drilling fluid circulation, hoisting, power supply, andblow-out prevention. These subsystems are designedto work seamlessly together. Figure 1 shows a typicalonshore drilling rig.

Current drilling practices are based on the princi-ple of rotary drilling, where a dedicated rotating sys-tem drives a cutting tool (i.e., drill bit) to break therock. A continuous weight is applied to the bit to pro-vide the necessary cutting force. The weight on thebit is controlled via the hoisting system, which is alsoused to lift and trip the well tubulars in and out of thewell. The cuttings are removed from downhole andtransported to surface by a drilling fluid (air, foam, ordrilling mud). Cuttings removal is the main functionof the circulating system that allows the drilling fluidto be pumped downhole (via the drill pipe) andreturned to surface (via the annular space betweendrill pipe and hole), where it is reconditioned.

Until relatively recently, the oil and gas drillingindustry portrayed an image using equipment andworking environment that was “large, dirty anddownright ugly.” This is unacceptable nowadays andthe industry pays close attention to making drillingrigs clean, safe, and environmentally friendly.Consequently, modern oil and gas drilling rigs havebecome ergonomic and tailored to the driller’s needfor a safe and adequate working space.

Modern Drilling Rigs

Modern drilling rigs rely on integrated mecha-nized equipment and ergonomic working space.Figure 2 shows an example of a highly computerizedcontrol system used to operate the modern rig, asopposed to the high level of labor required for a con-ventional rig.

A major feature of a drilling rig, particularly whenit is used in environmentally sensitive areas or urban

locations, is its footprint. The size of a rig’s footprintdepends on rig type (mobile vs. fixed) and well spec-ification (which is a function of target depth and geol-ogy). Figure 3 shows the overall footprint of a typicalonshore drilling rig.

The transition from traditional to modern drillingsites has required much closer collaboration betweendrilling engineers and architects. This has led to thedesign and construction of ergonomic and tailoreddrilling structures to meet the needs of today’s energyindustry. Examples of using modern rigs to drill inurban areas will be shown later in the article.

Drilling Deep Wells in Urban Areas

In some cases, wells must access reservoirs thatare situated directly underneath or in close proximityto urban areas. In other cases, wells must be drilled toinvestigate the geology of an area prior to building on

2 BULLETIN OF SCIENCE, TECHNOLOGY & SOCIETY / Month XXXX

Figure 1. A Typical Onshore Drilling Rig and Its MainComponents

Figure 2. Driller’s Cabin on a Modern Drilling Rig Comparedto a Driller’s Working Location in 1942Source: Reinicke and Teodoriu (2005).

or tunneling through the land. If the target depth is shal-low, the rig will not remain in place long enough to dis-rupt the local “way of life.” For example, an artesianwell with a shallow water table may only take a coupleof days to drill. However, for deeper wells, there aremany issues that cause concern when drilling in urbanareas, such as

• Risk of explosion• Risk of pollution of air and potable water• Noise level and vibrations• Impact on the landscape• Conformity with local environmental regulations• Space taken away from urban development• Length of time that the rig stays on site• Footprint of the drilling site

There are numerous examples of urbanization of thedrilling industry for hydrocarbon exploitation, whichcan be used as a reference for geothermal applications.

One is the THUMS Island project (OccidentalPetroleum Corporation, 2007; Oxy Long Beach Inc,2000) in Long Beach, California, which takes its namefrom the original consortium of Texaco, Humble,Unocal, Mobil, and Shell that operated the Wilmingtonfield. THUMS is a model example of how drilling oper-ations can be run in a sensitive environment.

It consists of four artificial islands, named Grissom,White, Chaffee, and Freeman, after the NASA astro-nauts who died in training accidents early in the U.S.space program. Built from boulders and sand, theymimic resort islands to blend in with the surroundingcoastal environment of Long Beach Harbor.

THUMS is the largest crude oil producer in the LosAngeles area. Despite the high rate levels of theTHUMS development, there has not been any pipelineleak since the beginning of production. Because of theunique successes that the THUMS project has achieved,it has already been awarded special recognition by localand national groups for environmental protection, com-munity beautification, and outstanding design.

Figure 4 shows the architectural aspects of rig cam-ouflage at THUMS.

Cases like THUMS make it possible to produce oiland gas without offending environmentally sensitiveplaces and without interfering with the urban andtourist development of inhabited places.

The following examples of wells drilled in urbanareas for geothermal energy recovery illustrate howand when it is possible for drilling rigs and urban lifeto coexist with one another.

There are two main types of geothermal energy: lowtemperature (less than 100°C) and high temperature(greater than 200°C). Although the high temperaturesources are used for electrical energy generation, the lowtemperature sources become more attractive when theyare coupled directly with on-site heat utilization (e.g.,heating of buildings, hot houses, etc). In the latter case, itis not always possible to keep the drilling rigs outside thecities, so architects and engineers must work together todesign the plant in conjunction with the infrastructurethat will directly use the geothermal energy. Although the daily running costs of generating electricity are rela-tively low, large initial capital investments are usuallynecessary. With the latest energy conversion techniques,

Teodoriu, Falcone / INCREASING GEOTHERMAL ENERGY DEMAND 3

Figure 3. Aerial View of a Drilling Rig FootprintSource: Reinicke and Teodoriu (2005).

Figure 4. Camouflaged Oil Rigs on Grissom Island, LongBeach, CaliforniaSource: California Department Of Conservation (2007).

geothermal resources may become more attractive forareas where low underground temperatures are found.One such example is represented by the modern conver-sion techniques based on the “Rankin and Kalima cycle,”which enables electric power generation at temperaturesas low as 100°C (WEA, 2004).

The following is a review of worldwide projects inwhich the geothermal energy from a deep well is usedto heat or cool buildings or districts. Projects of thistype can be classified as those involving new build-ings, for which the geothermal well is an integral partof the building (see Figure 5), or as projects that usedistrict heat transport infrastructures to connect thegeothermal facility (well plus heat exchangers) tobuildings. Note that the injection and production wellscan both be on location when directional drilling tech-niques are used.

When new buildings are designed, the well con-struction requirements are based on the heat capacityof the building. Well depths may vary, but the pro-duced fluids’ temperature range is typically between45 and 80°C. Existing projects, which used mobiledrilling rigs, reported that the major problems experi-enced were size of rig site, noise, and pollution(SuperC, 2006). Figure 6 shows the building of the so-called “SuperC” project in Germany, which usesgeothermal energy to supply heat to the new “Student-Service-Center SuperC” at RWTH Aachen University.The center will be heated using a deep well, “RWTH-1,” that is drilled adjacent to the RWTH main building(SuperC, 2006). The well temperature at 2,500 m is

about 85°C and the well is expected to deliver hotwater at a temperature of at least 70°C for the build-ing’s heating and cooling supply. It is a closed system,which means that it can be installed in any rock type.Because of the specific geology of the Aachen region,the chosen solution is very important as the hotsprings must be protected from any possible contami-nation.

According to some authors (SuperC, 2006), “theproject Super C and geothermal energy proves thatgeothermal heat supply for large buildings is techni-cally feasible and economically efficient, although theinitial costs are relatively high.”

The projects that use geothermal energy to heat orcool existing buildings or districts face more complicatedissues. As most of these projects are applied for largepublic buildings, often being tourist attractions, the rigsmust be part the architectural environment. Drilling andcompleting a well for heat mining can take up to severalmonths, but the area cannot be closed to the public or totourists for this length of time.

Another major project from Germany, shownschematically in Figure 7, is considering using a geot-hermal source to heat and cool the German parliamentusing a pair of wells; one shallow (for cooling) andone deep (for heating) (Bussmann, 2007).

The deep heat mining project to produce electricalpower and district heating in Basel, Switzerland, was ini-tiated in 1996, being partially financed by the FederalOffice of Energy and supported by private and publicinstitutions. The drilling target is located at a depth of

4 BULLETIN OF SCIENCE, TECHNOLOGY & SOCIETY / Month XXXX

2500 m

Heated rock

The “Super C” building

Geothermal well as heat exchanger

Figure 5. Schematic of a Well Used for Heat Mining as anIntegral Part of a New Building DesignSource: Modified after Sanner (2005).

Figure 6. The SuperC Project of the RWTH Aachen University,GermanySource: Superc (2006).

5 km (3.125 miles) below the surface, where theexpected average water temperature is about 200°C.Figure 8 shows a schematic diagram of the target welllocation according to the original plan. It must be notedthat the first attempt at drilling this geothermal wellfailed because of geological complications and overload-ing of the drilling rig (Haering & Hopkirk, 2002). A newrig with higher hook load capacity was purchased andthe well was successfully drilled at the second attempt.

Classification of Drilling Rigs Used to Drill in Urban Locations

The concept of modern drilling rigs was introducedearlier in this article. In what follows, a review of

some of the main modern rig designs used specificallyto drill wells in urban areas will be presented.

Conventional Drilling Rigs Used for UrbanDrilling Activities

For deep heat mining projects (e.g., drillingthrough hot dry rocks), the drilling rigs normallyadopted by the oil and gas industry have proved to bethe best option as they can drill quickly to the targetdepth. Unfortunately, the recent rise in oil and gasprices has caused an increase in demand for rigs, trig-gering a corresponding increase in their day rates,which has made it uneconomic to use oil and gasdrilling rigs for geothermal applications. Despite this,several services companies are designing their futurerig fleet with the geothermal drilling capability. Forexample, ITAG’s Rig 40 (Gutsche, 2005) shown inFigure 9, allows rig transportation and erection in amuch shorter time than for any equivalent oil rig. Therig was originally conceived for transportationthrough the narrow streets of southern Germany and

Teodoriu, Falcone / INCREASING GEOTHERMAL ENERGY DEMAND 5

20ºC70ºC

23ºC

Cold storage

Hot storage

5ºC

German Parliament

Figure 7. Geothermal Heating and Cooling of the GermanParliamentSource: Modified after Bussmann (2007).

Figure 9. The New Generation of Light Land Drilling Rig ThatCan Be Used for Both Oil and Gas Wells and Deep GeothermalWellsSource: Gutsche (2005). Copyright of ITAG Tiefbohr.

Figure 8. The Concept of the Deep Heat Mining ProjectSystem, Basel, Switzerland

northern France (see Figure 10), hence its modulardesign. The low footprint of Rig 40 (3500 m2) is com-parable to that of conventional mobile rigs, but itsmaximum hook load and mud tank capacity are muchhigher.

Figure 11 shows the conventional drilling rig usedto drill the deep geothermal well in Basel, discussedearlier. Although the well has proved to be a successfrom the point of view of recovering energy in anurban environment for district heating, the rig hascaused obvious disruption to the local architecture.The only way to reduce the negative impact on thelocal architecture is to reduce the height of the drillingrig to ensure an easy camouflage. Figure 12 shows theheight comparison between a conventional rig fordeep drilling and a modern or unconventional drillingrig, which considerably reduces the overhead workingspace requirement.

Unconventional Drilling Rigs Used for UrbanDrilling Activities

Any type of drilling rig that is not conforming tostate-of-the-art conventional systems can be classifiedas unconventional. As new technology becomes avail-able, unconventional rigs evolve further until today’sunconventional systems become the conventional rigsof tomorrow.

The current unconventional drilling rigs are charac-terized chiefly by new designs for the hoisting sys-tems to allow very high load capacity within smallstructures. Other key features of these rigs are:

• Electrical hydraulic drives• Ability to use the AC electric supply from a city

network• Low height (single derrick)• High load capacity• Partially or fully automated control systems• Modular construction

To the knowledge of the authors, only two companies,Drilltec and Herrenknecht Vertical, have publishedinformation about drilling rigs specifically designedfor deep geothermal applications. The main designfeatures that make these rigs appropriate for urbandrilling are briefly described below.

6 BULLETIN OF SCIENCE, TECHNOLOGY & SOCIETY / Month XXXX

Figure 10. Typical Small Roads Through French Villages

Figure 12. Conventional Drilling Rig (45 m to 55 m High, CanAccommodate Three Drill Pipes) on the Left Versus Unconven-tional Drilling Rig (15 m to 18 m High, Can Accommodate OneSingle Drill Pipe) on the RightNote: One single drill pipe is approximately 9 m long.

Figure 11. Conventional Drilling Rig Used for the DeepMining Project in Basel, Switzerland

The Drilltec Rig VDD370

The maximum hook load of 3.700 kN makes this rigsuitable for deep drilling applications (oil and gas andgeothermal). The major components of the DrilltecSystem (topdrive, pipehandling system) are hydrauli-cally driven. The power packs, that are electrically dri-ven, supply aggregates for hydraulic energy, provide the flexibility to use prime energy supply from a city network or—if there is not enough electrical poweravailable—a diesel motor-driven generator set (standardversion) can be used instead. Figure 13 shows theDrilltec rig and the driller’s cabin (Drilltec, 2006) fromwhere all operations can be controlled. Both, the smallfootprint and the low height of total 31 meters, advantagethe application of the super single rig in urban areas. Therack and pinion type rig is noise protected in many com-ponents. To absorb the sounds of the topdrive (equippedwith six hydraulic motors for traveling and fourhydraulic motors for rotation) the gears are covered com-pletely. In addition, most of the noise generating equip-ment (e.g., pumps, generators, power packs) is mountedin insulated containers.

The Herrenknecht Vertical Terra Invader 350 (TI-350) Rig

In response to the worldwide increase in drillingactivity and the lack of drilling capacity in the German

market, a new versatile drilling rig has been developedwith direct application to geothermal deep drillingprojects. The Federal Ministry for the Environment,Nature Conservation and Nuclear Safety has pro-moted the development of new a deep drilling rig aspart of its energy research program.

Being built for urban drilling activities, the new rig,shown in Figure 14, has a marked improvement innoise reduction. The main innovation of this drillingrig design is its hydraulically driven hoist system (lin-ear hydraulic cylinders) and noise insulated containersfor prime drivers and mud pumps. It has been reportedthat the expected noise pressure level at a distance ofapproximately 150 m from the rig is about the same asa domestic TV or radio at normal listening volume(Herrenknecht, 2006).

To achieve maximum drilling performance andsafety, the new rig design uses extensive integration ofthe automated processes (hands off technology).

Teodoriu, Falcone / INCREASING GEOTHERMAL ENERGY DEMAND 7

Figure 13. Drilltec’s Super Single Drilling Rig With AutomaticPipe Handling SystemSource: Drilltec (2006). Copyright of Drilltec GUT.

Figure 14. Herrenknecht Vertical Terra Invader 350 Drilling RigSource: Herrenknecht (2006). Copyright Herrenknecht.

After final testing, this new deep drilling rig will beemployed in the commercial zone of the municipalityof Duembaar (southern Germany). The planned geot-hermal power plant is expected to generate 5 to 6 MWof electric power and 25 MW thermal.

Future Needs of Drilling in UrbanEnvironments

To accommodate the future needs of drilling inurban environments, the new breed of drilling rigs willhave the following design characteristics:

• Modular construction for easier mobilization andtransportation through narrow city streets. Beingmodular will mean that most of the rig equipmentwill be mounted in small containers of small size,which has the added benefit of reducing noise emis-sion by sound insulation of the containers’ walls.

• Low noise emissions outside the drill site. One of themandatory changes to rig design to reduce noise levelsis to lower the rig’s height by using single drilling rigsinstead of triple rigs. Noise reduction is also obtained byusing unconventional hoist systems (e.g., rack and pin-ion system, linear hydraulic motors)

• High hook load capacity, despite the small size of thedrilling rig, which can be achieved by means of uncon-ventional technologies

• Easily camouflaged for long-term drilling projects• Smart power supply, using on-site diesel generators or

tapping into the city electrical power grid, and havinglow emissions of noise, vibrations, and exhaust gases

• Closed-loop spill and mud processing unit to avoidany contamination of the rig site

• Small size of drilling rig, which will be achievedthrough integrating the automated processes and com-puter controlled rig floor equipment

Conclusions

The urban environment forces rig design to evolveto accommodate in-city drilling operations.

A review of international projects that aim to integratedrilling with urban infrastructures has been presented inthis article. The new trends in harmonizing urban archi-tecture with drilling rig design have been discussed.

With a rising number of the geothermal deep heatmining projects around the world, there is a slowlyincreasing need for drilling rigs to operate in an urbanenvironment, with all the problems that entails.

Although the market is small in comparison with theoil and gas industry, the future energy trends may radi-cally change the demand for “urbanized drilling rigs.”

References

American Ground Water Trust. (2007). How wells are drilled.Retrieved March 15, 2007, from http://www.agwt.org/info/pdfs/welldrilling.pdf

Arabas, I. (2005). The founder of the Polish and world oil industry:Polish pharmacists Ignacy Lukasiewicz and the 150th anniversaryof the lighting of the first kerosene lamp. Pharmaceutical History,35(1), 8-10.

Baker, R. (1996). A primer of oilwell drilling (5th ed.). Austin, TX:Petroleum Extension Service and International Association ofDrilling Contractors.

Bussmann, W. (2007). German geothermal trends and developments.Retrieved March 15, 2007, from http://www.geothermie.de

California Department of Conservation. (2007). Home. RetrievedMarch 15, 2007, from http://conservation.ca.gov/

Drilltec. (2006). Large scale drillings—Worldwide. RetrievedDecember 15, 2006, from http://www.Drilltec.de

Gutsche, W. (2005, June). A new generation of AC-driven light landrigs. Paper presented at the Institute of Petroleum Engineering,TU Clausthal, Clausthal, Germany.

Haering, M., & Hopkirk, R. (2002, March). The Swiss deep heat min-ing project: The Basel exploration drilling. Paper presented at theproceedings of the International Summer School Conference:International Geothermal Days, Germany 2001, Bad Urach,Germany.

Herrenknecht. (2006). Technology going deep. RetrievedDecember 15, 2006, from http://www.herrenknecht-vertical.de

Lungonelli, M., & Migliorini, M. (2003). Piero ginori conti.Laterza, Italy: Collana Cultura e Industria.

Occidental Petroleum Corporation. (2007). California-THUMS.Retrieved March 15, 2007, from http://www.oxy.com/OIL_GAS/world_ops/usa/thums.htm

Oxy Long Beach Inc. (2000, December). THUMS Long Beach company overview. Presented at the West Coast PTTC Workshop,Los Angeles.

Reinicke, K. M., & Teodoriu, C. (2005). Neue entwicklungen in derbohrtechnik. Heft, Germany: Akademie der Geowissenschaften zuHannover Veroefentlichungen.

Sanner, B. (2005, April). Geothermal energy—Country profile for Germany. Workshop on Regulatory and Economic ToolsGoverning the Enhanced Exploitation of Geothermal Energyin the European Union, Kistelek, Hungary.

SuperC. (2006). The geoscientific research. Retrieved December15, 2006, from http://www.superc.rwth-aachen.de

Talwan, M., Belopolsky, A., & Berry, D. L. (1998). Geology andpetroleum potential of central Asia. Houston, TX: RiceUniversity, Baker Institute.

U.S. Department of Energy. (2006). A history of geothermal energy in the United States: Energy efficiency and renewable energy.Retrieved December 15, 2006, from http://www1.eere.energy.gov/geothermal/history.html

World Energy Assessment. (2004). Energy and the challenge ofsustainability. New York: Author.

Catalin Teodoriu is an assistant professor in the Harold VanceDepartment of Petroleum Engineering at Texas A&MUniversity. He has an equivalent MS (1996) from “Oil and

8 BULLETIN OF SCIENCE, TECHNOLOGY & SOCIETY / Month XXXX

Gas” University of Ploiesti and two PhDs (2003, TechnicalUniversity of Clausthal; 2005, “Oil and Gas” University ofPloiesti). He has worked as a R&D engineer for 2 years withR&D Center of Petrom (actually OMVRomania) Company,more than 6 years as researcher, and later on as researchsupervisor and equivalent associate professor with the techni-cal University of Clausthal. At technical University ofClausthal, he was involved in teaching graduate and under-graduate drilling and completion courses and in drilling, wellintegrity, and drilling fluids research for more than 5 years. AtTexas A&M University, he is involved in teaching undergrad-uate drilling courses and in drilling and well integrityresearch. Related research activities that he has been involvedwith include casing resistance under extreme loads, swellingcements for gas wells, drillstring components makeup proce-dures, underbalanced drilling and formation damage, evalua-tion of the casing fatigue and fatigue of casing connectors, anddevelopment of laboratory testing devices and facilities.

Gioia Falcone is an assistant professor in the Harold VanceDepartment of Petroleum Engineering at Texas A&MUniversity. She holds a first degree in petroleum engineer-ing from the University “La Sapienza” of Rome, an MScdegree in petroleum engineering from Imperial CollegeLondon, and a PhD in chemical engineering from ImperialCollege London. Prior to joining Texas A&M, she workedwith ENI-Agip, Enterprise Oil UK, Shell E&P UK, andTOTAL E&P UK, covering both offshore and onshoreassignments. She has served on the SPE ATCE TechnicalProgramme Committee and, on several occasions, on theSPE ATCE Well Operations Subcommittee. She is a techni-cal editor for the SPE Projects, Facilities & ConstructionJournal, memberships chairperson for the SPE Researchand Development Technical Section, and a member of theEditorial Advisory Board for Scientific Journals Interna-tional. Her main research interests are in the areas of inte-grated production systems and multiphase flow.

Teodoriu, Falcone / INCREASING GEOTHERMAL ENERGY DEMAND 9

SPE 122786-PP

Design of a High-Pressure Research Flow Loop for the Experimental Investigation of Liquid Loading in Gas Wells J. Fernandez(*), SPE, G. Falcone, SPE and C. Teodoriu, SPE, Texas A&M University

Copyright 2009, Society of Petroleum Engineers This paper was prepared for presentation at the 2009 SPE Latin American and Caribbean Petroleum Engineering Conference held in Cartagena, Colombia, 31 May–3 June 2009. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract Existing models to predict and analyze liquid loading in gas wells are based on steady-state flow. Even when transient multiphase wellbore models are employed, steady-state or pseudo steady-state inflow performance relationships are used to characterize the reservoir. A more reliable approach consists of modeling the dynamics in the near-wellbore region with its transient boundary conditions for the wellbore. The development of new models to mimic these dynamics requires a purpose-built flow loop. We have developed a design to construct such a facility.

This new facility will be the first to integrate pipe representing the wellbore with a porous medium that will fully mimic the formation surrounding the wellbore. This design will account not only for flow into the wellbore, but any reverse flow from the pipe into the medium.

We used integrated wellbore/reservoir system analysis to screen the parameters required to recreate liquid loading under laboratory conditions. Once the range in operating conditions was defined, the equipment and mechanical components for the facility were selected and designed.

Our results showed that three reciprocating compressors working in parallel provide the smallest, most economic, and most flexible configuration for the Tower Lab facility at Texas A&M University. The design of the pressure vessel hosting the porous medium will require a cylindrical body with top- and bottom-welded flathead covers with multiple openings to minimize weight. The required superficial velocities for air and water indicate that the system will need independent injection into the porous medium through two manifolds. Optimally, the system will use digital pressure gauges, coriolis or vortex technology to measure air flow and turbine meters for water flow.

The new facility will significantly improve our ability to mimic the physics of multiple phase flow for the development of liquid loading models and lead to better optimization of gas fields. Introduction Natural gas is one of the principle sources of energy for many of our day-to-day needs and activities. It is a vital component of the world's supply of energy as it is one of the cleanest, safest, and most useful of all energy sources. Strong domestic and international demand has made it commercially attractive since the 1950s (Yamamoto and Christiansen, 1999). However, the deliverability from existing gas reservoirs continues to decline and the scarcity of promising prospects has increased the need to maximize gas recovery from every gas well.

As natural gas is produced from a reservoir, the simultaneous flow of gas, liquid hydrocarbons, and water is a common situation for both land and offshore production systems (Toma et al., 2006). In vertical wells, water and intermediate hydrocarbons can condense in liquids in the wellbore, depending on the current composition of the gas during production. In depletion-drive reservoirs, the energy available to remove the produced/accumulated liquids, condensates, and formation water to the surface declines. This energy eventually becomes so low that flow rates decrease below a certain critical velocity, and fluids produced with the gas flow stream begin to fall back, no longer being carried to the surface. Where the flow stream is insufficient to continuously transport the well liquids upward in the wellbore, a wellbore liquid loading condition begins.

Loading up of liquid in the wellbore has been recognized as one of the most severe problems in gas production. Accurate prediction is of critical importance to efficiently take the optimal approach to deal with this situation. Although several investigators (Turner et al., 1969; Coleman and McCurdy, 1991; Guo et al., 2005; Dousi et al., 2005; Gool and Currie, 2007)

have suggested methods to predict the onset of liquid loading, results from these models are only applicable to steady-state flow; however, liquid loading corresponds to unsteady-state flow conditions, both in the well and in the near-wellbore region of the reservoir.

(*) Now with Anadarko Petroleum Corporation

2 SPE 122786-PP

A more reliable approach would be to use a transient multiphase flow wellbore model that accounts for the dynamics in the near-wellbore region via transient boundary conditions. A dynamic, integrated (reservoir/wellbore) system modeling approach would predict conditions for transition from an acceptable production flow regime (annular or mist flow) to an unacceptable regime (churn flow) that triggers liquid loading in the wellbore (Solomon and Fernandez, 2007). Achieving these objectives, requires rigorous empirical studies of transient multiphase flow under laboratory conditions that attempt to mimic well behavior under liquid loading.

A review of existing flow loops worldwide revealed some specialized areas of research such as liquid loading in gas wells are still lacking dedicated test facilities (Falcone et al., 2007). Having an experimental facility with adequate vertical height and diameter of the test section to reproduce the flow regimes encountered prior to and after onset of liquid loading in gas wells is necessary to conduct experimental studies to mimic the dynamic boundary conditions that exist between the reservoir and wellbore during liquid loading. This requires a compression and boosting system capable of circulating fluids at flow rates, pressures and high gas volume fractions (GVF) similar to real gas-field production situations, as well as, technology and instrumentation for controlling and monitoring key two-phase flow parameters such as pressure, temperature, and flow rates among others.

Additionally, the facility will allow the creation of the flow resistance represented by the reservoir, for which the common practice has been to use a series of valves to simulate permeability variations. However, flow from a real reservoir to the wellbore goes through a porous medium, and in the case of multiphase flow, relative permeability must be taken into consideration.

To recreate the effect of the near-wellbore region, our new facility (TAMU Tower Lab) will allow us to attach a cylindrical pressure vessel containing tightly packed spherical glass beads to the base of a vertical multiphase flow loop. Flow of fluids into the porous medium will be provided by entry points symmetrically aligned at the pressure vessel using an injection manifold with adequate geometry to ensure proper dispersion of the fluids. Liquid Loading in Gas Wells The process of liquid loading can be summarized in four stages (Neves and Brimhall, 1989):

1. After the well has been completed and production begins, a gas well has enough energy from its high initial reservoir pressure and gas flow rates, to carry the liquids all the way to the surface (Fig. 1(a)). At this stage the gas velocity is greater than or equal to the critical velocity required to continuously remove the liquids in the gas stream.

2. As production continues, natural reservoir pressure declines, producing a decrease in gas flow rate that directly induces a decrease in gas velocity until reaching and falling below the critical gas velocity value. Consequently, liquid droplets suspended in the gaseous phase will begin to move downward and start accumulating at the wellbore, restricting the effective flow area for the gas and impeding its continuous production (Fig. 1(b)).

(a) (b) (c) (d)

Fig. 1 – During gas production, flow regime changes from mist (a) to annular (b) to slug/churn flow (c),

which causes the well to load up and die

3. Followed by the accumulation of the liquids, gas flow rate initially increases because of the reduction of the effective area, which results in a larger pressure drop across the accumulated liquids downhole. The pressure drop increases until the downstream pressure reaches the pressure necessary to transport the liquids up the tubing (Fig. 1(c)).

4. As a well cycles back and forth between the last two stages, the time differential between produced liquid slugs at the surface becomes greater as a consequence of the time required by the reservoir to reach a pressure high enough to blow the liquid slugs up the string. Eventually, the backpressure at the sand face originated by the liquids that have accumulated at the bottom overcomes the available reservoir energy, causing the well to load up and die (Fig. 1(d)).

SPE 122786-PP 3

The well-established techniques to alleviate the effects of liquid loading have been described by several authors (Stephenson et al., 2000; Lea et al., 2003). Foaming the liquid water can enable the gas to lift water from the well. Using smaller tubing or creating a lower wellhead pressure sometimes can extend mist flow. The well can be unloaded by gas lifting or pumping the liquids out of the well. Heating the wellbore can prevent liquid condensation. Downhole injection of water into an underlying disposal zone could be another option. Each of these technologies has its niche, but choosing the optimal strategy depends on a collection of factors including composition of reservoir fluid, operating pressures and temperatures, and of course economics (Yamamoto and Christiansen, 1999). Comparison of Major Equipment and Operating Conditions of Multiphase Research Flow Loops Table 1 illustrates the maximum pressure and the equipment used to handle the liquid and gas phase for the flow loops described in this investigation.

Table 1 – Major equipment used in research loops for liquid injection and gas compression

Flow loop Maximum Pressure,

psi Liquid phase Gaseous phase

South West RI (Hart, 2007) 3,600 Centrifugal pumps Centrifugal compressor NORSK HYDRO (Robole, 2006) 1,595 Centrifugal pumps + twin screw pump (for oil) Centrifugal compressor CEESI (Kegel and Kinney) 1,500 Positive displacement pump 4 stage Recips + 4 single stage recips SINTEF LSL (Unander, 2007) 1,305 Centrifugal pump Reciprocating compressor TUV NEL (www.tuvnel.com) 910 Centrifugal pump Centrifugal gas blower Boussens (Corteville et al., 1983) 725 Centrifugal pumps Centrifugal compressor IFP (Vilagines and Hall, 2003) 725 Two-phase pump IFE (Langsholt, 2007) 145 Centrifugal + dosage pumps + screw pump Multiphase pump (twin screw) SINTEF MSL (Unander, 2007) 117 Centrifugal pumps Multiphase pump (twin screw) LOTUS (Falcone et al., 2003) 90 Centrifugal pumps Compressed tank TAMU Tower Lab 120 Centrifugal pump Reciprocating compressor

Fig. 2 shows the vertical height for the multiphase research flow loops mentioned in Table 1. The maximum length of a flow

loop affects the development of different flow regimes, particularly in transient flow investigations (Falcone et al., 2007). The vertical height of the test section available of our design clearly is competitive when compared to other facilities.

The range of flow regimes that can be reproduced in a flow loop is related to the flow rates that can be circulated in the system. The maximum reported flow rates for gas in the flow loops identified for this study are given in Fig. 3. The highlighted portions of the graphs show that the proposed range in pressures and rates in our facility would cover a region where current research flow loops do not operate.

Fig. 2 – Vertical length vs. Pressure

Fig. 3 – Vertical length vs. Max. Gas rate

Results of the extensive review indicate a niche in operating conditions for the TAMU Tower Lab; however, the final

values for an experimental facility have to be selected for the type of research to be conducted. Since our facility is intended for the experimental investigation of liquid loading, values of pressure and flow rates including adequate geometry and hardware should allow us to mimic the conditions that precede and follow liquid loading in a gas well. General Considerations for Flow Loop Design From a research standpoint the modeling of hydrocarbon processes in a laboratory environment and the existing operating conditions encountered in real oil and gas fields is significant. However, this breach can be narrowed by designing a state-of-

4 SPE 122786-PP

the-art facility operating under more realistic conditions of pressure, flow rates, and temperatures found in real gas wells. Nevertheless, the exact representation of an existing field is difficult to accomplish entirely.

Flexibility is an important characteristic the multiphase flow loop should incorporate at an early stage of the design; however, as a result of the high capital investment needed to build a facility to study multiple problems within the area of multiphase flow, making a flow loop 100% flexible is neither technical nor economically feasible. The approach by most laboratories is to design and build their facilities to accommodate a range in operation and fluids to be handled by their flow loops.

Current models used to study multiphase flow have been obtained and validated via research flow loops. However, new investigations are continuously being carried to out to improve existing correlations or to propose new methods. This creates a need for more problem-oriented research flow loops.

For our facility, we identified the range in operating conditions through a review of multiphase flow loops around the world. We validated these values throughout this study to guarantee that conditions that precede and follow liquid loading can be achieved. Fig. 4 shows the proposed downscaling of a gas well completion for our facility.

Fig. 4 – Our laboratory will introduce a porous medium to better capture flow behavior in the near wellbore region

The assembly will attempt to mimic the dynamics of the near-wellbore region to provide boundary conditions for the development of predictive transient multiphase flow models.

To recreate the pressure effect of the near-wellbore region we will attach a cylindrical pressure vessel containing tightly packed spherical glass beads to the base of the vertical multiphase flow loop. Glass beads, which represent a formation of uniform grain size, are low cost and readily available, and their use in fluid flow experiments is well documented (Alshuraiqui et al., 2003, Costantini, 2005).

Average porosity of a glass bead pack is in the order of 38% with permeability ranging from 1.2 to 29 Darcy, depending on their size and packing. Fig. 5 is a front view section of the pressure vessel containing glass beads tightly packed.

Fig. 5 – Pressure vessel will recreate effects of the near-wellbore region

Flow of fluids into the porous medium may be provided by entry points symmetrically aligned at the pressure vessel by means of a distribution mechanism that ensures proper dispersion of the fluids. Fig. 6 illustrates the use of a single injection manifold where the testing fluids would mix prior their arrival at the pressure vessel.

Production tubing

Packer

Gas enters through perforations

Production casing

Cement

Surface casing

Perforations

Gas produced to surface through tubing

Choke

Manifold

Production tubing

Edge of reservoir

Porous medium

Air Water

SPE 122786-PP 5

Fig. 6 – A single injection manifold will ensure proper mix of fluids entering the system The phases initially considered as testing fluids are air and water. This will demand using a compression and boosting

system capable of delivering the fluids at the required operating conditions. Control and monitoring will be managed by flow or mass meters. During transient flow periods, measurements of void fraction, temperatures and pressures will be taken at different locations across the flow loop.

Design factors for mimicking liquid loading at TAMU Tower Lab

Solomon and Fernandez (2007) performed a sensitivity analysis using a commercial wellbore/reservoir simulator and a response surface methodology to minimize cost and investigate different design parameter settings for the facility. Table 2 presents the design variables investigated, which included tubing diameter sizes, glass beads’ packing permeability, cylindrical pressure vessel dimensions, and operating pressures.

Table 2 – Design factors (Solomon 2007)

Factor Numerical range Permeability, Darcy 0.01 to 6,000 Tubing diameter, in. 2.72 to 4.5 Compressor discharge pressure, psia 470 to 650 Pressure vessel radius, ft 1 to 6 Pressure vessel height, ft 0.7 to 3 Water Gas ratio, STB/MMscf (constant) 3,300 Surface choke pressure, psia 20 to 400

Published literature suggests that the typical diameter of glass beads used to perform fluid flow experiments typically ranges from 250 μm to 3mm, with permeability and porosity values of 29 to 1,200 Darcy and 38 to 46%, respectively. However, we will conduct laboratory experiments with glass beads to determine values of permeability and porosities that are apropriate for the artificial porous medium.

Tubing size diameters must be typical for gas-producing wells, which required analysis of the effect of the deliverability of the system, considering other variables. The range given in Table 2 considers diameters found in most gas well completions in the US.

A compressed air system, providing a discharge pressure between 470 to 650 psi with gas rates between 400 to 650 scf/m, was considered necessary to evaluate the impact on the flow potential of the system. We plan to use a compressor capable of delivering operating conditions that can mimick liquid loading. In addition, the analysis assumes a constant water gas ratio of 3,300 STB/MMscf and takes into consideration the operating conditions (pressure and flowrates) that exist at the nozzle ports located on the external entry points of the pressurized vessel. Fig. 7 shows the location of the four nozzle ports.

6 SPE 122786-PP

Fig. 7 – Operating conditions are supplied at the 4 nozzle ports surrounding the pressure vessel

For the design of the pressure vessel we plan to make optimal use of the available area in the basement (56 ft2). The

dimensions of the vessel must guarantee safe operations with adequate space for the installation of flow metering and pressure monitoring devices. As for the surface chokes, we investigated a range of 20 to 400 psi.

A multiple response optimization (Solomon and Fernandez, 2007) identified options to minimize cost and maximize system performance. The recommended design factor combination was to use 3 in. tubing with a 500 psi compressor and cylindrical pressure vessel dimensions of radius 4.85 ft and height of 2 ft. System calculations using the wellbore/reservoir simulator with these parameters validated their applicability to simulate liquid loading in our facility. Our results from generating IPR and VLP curves confirmed the feasibility of using the parameters considered in the optimization process. Design Option for the Boosting System Three possible arrangements for compression and boosting include a compressor with a water pump, a multiphase pump, or a pressure vessel storing compressed air. Our analysis showed that the pressure vessel option was not feasible, as it would not be capable of delivering the required flow rate and pressure for the time required to run the experiments. The compressor plus pump option proved to be cheaper than the multiphase pump.

Fig. 8 illustrates the system where a compressor and pump deliver independent or commingled injection at the manifold. In addition, two manifolds allow flexibility for single fluid injection or for mixture of the phases prior to entering the bottom section of the tubing.

Fig. 8 – Compressor and water pump will allow independent or commingled injection of fluids

This design is practical, and it allows for both single and multiphase flow through the system. Additionally, cost reductions

are achieved by eliminating the use of a multiphase pump, which costs more than any other pump and also imposes a challenge from a system-control standpoint. An initial stage of compression given by a small unit, followed by a booster (such as a multiphase pump) would add higher control complexity to the overall compressed air and water system as compared to the

4X Nozzle ports

Single/multiple stage compression

Pump

T-joint

Water manifold

Non-return valve

Valve Bottom section of tubing

Meter

Meter Non-return valve

Air manifold

Two-phase manifold

Valve

Valve Valve

SPE 122786-PP 7

system shown in Fig. 8. Furthermore, our design option involves less equipment and so minimizes maintenance complexity and cost. Although larger equipment may be desirable to provide the complete operational envelope, it costs more and must fit into the physical space available. Facility Design The approach to designing the facility that will accommodate all the major equipment and components for the operation of the multiphase flow loop at Tower Lab began by establishing the operating conditions. Since one of the main challenges of this project was to properly size the compressed air system, our design set out with the design of the pressure vessel, then progressed upstream, through the injection manifold, pipe lines, instrumentation and compressed air package. Flow Loop The pressure vessel will be located at the basement of the Joe C. Richardson Building on the Texas A&M University campus. The first task was to identify the space available for the proper installation and operation of the equipment. Fig 9(a) is a top view of the building basement, and Fig. 9(b) highlights the possible location for the pressure vessel.

Fig. 9 – Basement at Richardson building and location of pressure vessel

However, it is important to mention that the area in the basement does not extend throughout the height of the building

therefore, measurements of every floor were taken to guarantee a free path for the tubing. Fig. 10 is a top view of the Tower Lab from 2nd to10th floor.

Fig. 10 – Tower Lab main dimensions from 2nd to 10th floor

Three openings can be allocated to support the vertical tubing; however, the central opening is used by a crane, which is installed at the roof of the building. This left only two possible locations for the pressure vessel and of these, the one closest to the wall had to be eliminated as it would limit the radius of the vessel. As Fig. 10 shows, these openings have “I” beams installed at both sides (on every floor), which are useful for holding and maintaining the verticality of the pipe sections.

Pressure Vessel By placing the tubing as indicated in Fig. 11, and having the geometrical center shared by both the pressure vessel and the tubing, we used the maximum internal diameter of the container to optimize its location with respect to the space available, at every floor and in the basement.

Crane path

I-beams

8 SPE 122786-PP

Fig. 11 – Proposed location of pressure vessel

Geometrical parameters (Solomon and Fernandez, 2007) indicated an internal radius of 5 ft (60 in.) and uniform height of 3

ft (36 in.); however, in our facility the internal diameter had to be modified because of lack of space. We designed the pressure vessel providing the largest diameter possible within the physical constraints imposed by the system.

We used ASME Code (2007) to design the different parts, including the shell, unstayed flat heads and covers, openings, reinforcements, nozzles, fittings, bolted connections, lifting lugs and support legs. The fabrication materials are also covered by ASME Code for pressure vessels. Table 3 presents the chosen materials as a function of operating temperature.

Table 3 – Material selection guide

Design Temperature ºF Material Plate Pipe Forgings Bolting/Nuts

33 to 775 Carbon Steel SA-285C SA-106-B SA-105 SA-193-B7 with SA-194-2H

We do not anticipate corrosion issues in our facility. We chose carbon steel as the fabrication material; however, as

experience has demonstrated, a fully non-corrosive environment is unlikely to be encountered in any process. As a consequence, we added an additional value of 0.125 in. to the required thicknesses to account for corrosion effects.

Several designs were technically and economically evaluated to not only minimize the fabrication costs, but also to facilitate the installation of the equipment. We used a commercial design software to generate different configurations to determine the best option in terms of expected fabrication costs, total weight, and (most importantly) the ease with which experiments could be conducted.

We evaluated having a flat head attached with bolts (options 1, 2, 3 and 4) and openings (options 3 and 4), a flanged and dished plate on the bottom (options 2, 4 and 6), and welded flat heads on top and bottom (options 5 and 6) with openings on top. Fig. 12 shows the different alternatives.

Fig. 12 – Configurations for weight reduction

Tubing attached to I-beam

Flanged & dished (Bottom head)

Flanged & dished (Bottom head)

Flanged & dished (Bottom head)

Our design

SPE 122786-PP 9

The chosen option offered outstanding weight reduction, but more importantly, the thickness of the porous medium was kept constant throughout the vessel radius. Additionally, we incorporated leg supports using 4 x 4 x 0.375 in. angle beams of structural steel and 0.625 in. thick base plates. The support system can withstand up to 10 tons, with additional lifting lugs designed to lift up to 14 tons.

Compressed Air System After identifying the capacity and pressure requirements, we performed an initial screening to select possible technologies for the compressed air system in our facility. Fig. 13 shows the results.

Fig. 13 – Screening of compressor types

According to the literature, air compression is possible using a centrifugal (surge problems might occur with varying flows), reciprocating or screw compressor. However, Snow (2003) showed that a screw compressor by itself is not capable of delivering the discharge pressure required by our system, so one option may be installing a booster to supply the additional increase at the discharge. The proposed solutions are summarized in Fig. 14.

Fig. 14 – Possible solutions for the compressed air system using single or multiple units

10 SPE 122786-PP

The initial approach for this project was to select the appropriate technology (i.e. compressor type) to meet the required

operating conditions for the Tower Lab. We contacted 22 manufacturers to get proposals for the facility from which only six qualified in terms of capacity and size. Some of the proposals involved using single or multiple units, all delivering different values of pressure and flow rates within our operating conditions range. Fig. 15 shows the different combinations. Compressor A

Compressor B

Compressor C

Compressor D

Fig. 15 – Compressor arrangements by different manufacturers

We used a scoring method known as the analytical hierarchy process (Saaty, 1977) to compare not only operating costs but also maintenance, initial and installation costs, footprint, flexibility, foundation requirements and service support (see Fig. 16).

Compressor E

Compressor F

SPE 122786-PP 11

Compressor supplier

MaintenanceService supportFootprint FoundationInitial

costFlexibilityOperating

costInstallation

Kaeser Cameron Atlas Copco Ingersoll Rand Sullair CompAir

Fig. 16 – Decision hierarchy for choosing a compressed air system supplier

Our analysis showed that Compressor E is the most attractive candidate because of the flexibility of having three units

working in parallel, each one providing high pressures and, when powered simultaneously, providing the required flowrate. No special foundations are necessary for the installation of the equipment and the initial operating and maintenance costs are competitive with the other candidates. In addition, the service manufacturers provide adequate support. (Fig. 17)

Fig. 17 – Analytic hierarchy process identified Compressor E as the best solution for our facility

Injection Manifold and Piping Design

The manifold is the element in charge of distributing the fluids provided by the compressed air and water systems to the artificial porous medium inside the pressure vessel. It is connected to the container through four equally-spaced entry points to propagate a radial-distributed pressure and flow. During the design phase, we encountered problems as a consequence of the space available to locate the vessel and surrounding it (Fig. 18).

Fig. 18 – Problems with initial design Alternatively, the options presented in Fig. 19 allow an adequate installation. Option (a) includes two straight segments of

pipe with a circular section, while option (b) has five straight pipes (b), yet both provide uninterrupted flow with lower pressure drops than a system with elbows and tees. However, we need to evaluate the feasibility of their fabrication with pipe

Point of higher pressure and induced flow rate

Not enough space

Compressor A Compressor C Compressor F Compressor E Compressor B Compressor D

12 SPE 122786-PP

manufacturers.

a) b)

Fig. 19 – Recommended manifold designs

The injection manifold required selecting an adequate internal diameter for the distribution of both air and water at the

entry points. Choosing a line size depends on both pressure drop and flow velocity. The initial design considered a single manifold, but evaluation of the internal diameter, based on the superficial velocities required for both phase,s dictated the need for independent injection. The water and air superficial velocities were obtained from our sensitivity analysis (Solomon and Fernandez, 2007) where parameters such as permeability, tubing size, pressure (at injection points), vessel radius, choke, and water-gas ratio were varied to identify recommended ranges of operation for different scenarios.

Table 5 presents the results for the range in internal diameter required for the water and air pipelines. The difference between sizes for the air and water lines suggested that independent injection, as shown in Fig. 20, was necessary to comply with the required superficial velocities. However, to implement a single manifold design, we will need more investigations on how air and water would behave in a shared pipe in order to account for the actual phase distribution in the pipe.

Table 5 – Pipe internal diameter range for air and water

Fig. 20 – Independent injection manifolds for two configurations

We used ANSI B31.3 (2002) to determine the wall thickness of the pipes considering the internal pressure, corrosion and material properties. We selected 1.5 in. and 3 in. nominal pipe size with 1.5 in. and 2.9 in. internal diameter for the air and water lines respectively.

Pressure drops in piping system and manifolds We calculated pressure drops in the water line system using Darcy’s equation Eq. 1. Darcy states that the friction head loss between two points in a completely filled, circular cross section pipe is proportional to the velocity head and the length of pipe and inversely proportional to the pipe diameter.

2

2LfLVH

gd= …………………………………………………………………………………………………………………..(1)

Pipe diameter range, in. Fluid Min. Max.

Air 0.532 1.660 Water 2.844 3.030

SPE 122786-PP 13

In most production facility piping systems, the head differences caused by elevation and velocity changes between two

points can be neglected. We finally calculated the pressure drop in the water line using equation 2.

26

511 5 10 lfLQp .

dγ−Δ = × ………………………………………………………………………………………………………(2)

The estimated distance for the piping network connecting both the compressed air and water system to the pressure vessel is approximately 188 and 190 ft respectively. A schematic of the proposed piping system for water and air flow is depicted in Fig. 21.

Fig. 21 – Schematic of piping system

Our approach to size the lines was the same as for the manifold. We propose to use 1.5 in. pipe for the two lines allowing for a change equal to 10 times the line diameter of the 3 in. water manifold. The water supply will be provided by a high pressure water pump located in the same area as the compressor (Fig. 22).

Fig. 22 – Water supply system

For the water line, the total length of straight pipeline is approximately 157 ft, but the minor losses from the installation of 8 long radius elbows increase the total equivalent length to 190 ft. Thus, the total pressure drop in the water line, including the manifold, is ~ 2.6 psi.

For gas pipelines, several empirical equations have been developed, using various coefficients and exponents to account for

efficiency and friction factors. We used the Darcy-Weisbach general gas flow equation to calculate pressure drops in the gas line, considering less than a 10% change in the inlet pressure.

Compressed air system

Water system

Air pipeline

Water manifold

Water supply

14 SPE 122786-PP

2

51

12 6 gSfLq ZTp .

p dΔ = …………………………………………………………………………………………………………(8)

The total length in straight pipes is expected to be 170 ft, increasing to 1885 ft with the installation of long elbows.

Additionally, a 1 in. enlargement will be located at the exit of the compressed air system, but the pressure drop at this point can be considered negligible. The final estimated pressure drop is ~ 1.685 psia. (Fig. 22)

Fig. 23 – Air supply system Monitoring and Instrumentation – Pressure gauges and Flowmeters The installation of pressure gauges in a research flow loop is of vital importance, not only to monitor the pressure at which experiments are conducted, but also to evaluate and assess pressure drops in the system and equipment performance. Table 6 presents the location and number of gauges to be installed at the Tower Lab:

Table 6 – Location and Quantity of Pressure Gauges for Tower Lab Location Quantity Air line prior arrival at injection manifold 1 Entry points of injection manifold 4 Water line prior arrival at injection manifold 1 Top surface of pressure vessel 4

Since the accuracy of most pressure gauges is better in the middle portion of a gauge, the range should be about twice the

maximum anticipated pressure. Furthermore, the maximum operating pressure should not exceed 80% of the full pressure range of the gauge. Using the maximum operating pressure of 580 psi given by the Compressor E system, the range for the pressure gauge should be 0 – 1,000 psi. The pressure gauges to be selected for the Tower Lab will be digital, and the accuracy should be at least be 0.25%, but we intend to evaluate the installation of gauges with 0.1 or 0.05% accuracy if costs are comparable. To this end, we reviewed initial costs and accuracy for 0.125 in. socket size, battery charged gauges with capacities from 0-1000 psi and the results are shown in Fig. 24.

Fig. 24 – Cost of Pressure gauges to be installed at Tower Lab will range from USD 750 to USD 1,500

As expected, the higher the accuracy, the higher the cost, so if we decide to install gauges with 0.05% accuracy we are looking at an initial cost of USD 750 to USD 1,500. In addition, gauges dimensions will be considered due to the limited space

Air manifold Air supply

SPE 122786-PP 15

on fittings located on top of the vessel. An example of installation of a model gauge with accuracy of 0.05% on the vessel shows no major issues with space availability is shown in Fig. 25.

Fig. 25 – Installation of model gauge on pressure vessel

Flow measurement is another “need to know” process parameter besides temperature and pressure, so accurate measurement of both air and water is critical in the operation of a research flow loop. Numerous types of flowmeters are available for closed-piping systems and these can be classified as differential pressure, positive displacement, velocity, and mass meters. Furthermore, they can be grouped into general categories, some of which may overlap with one another, but are still useful in describing some of the factors involved in flowmeter selection. These categories are: I) Flowmeters with wetted moving parts, II) Flowmeters with no wetted parts, III) Obstructionless flowmeters and IV) Flowmeters with sensors mounted external to the pipe.

Flowmeters selection is generally a process of elimination based on technical criteria such as pressure, temperature, specific gravity or density, viscosity and flow range, but specific technologies have been recognized as more appropriate for research facilities. Literature shows a tendency for using coriolis, vortex, and turbine meters. In particular, the coriolis technique has often been described as the near-perfect approach to for gas measurement, but at an average selling price of between USD 5,000 and USD 6,000, it is relativelyI more expensive technology. Table 7 presents a general comparison between the different technologies used in metering.

Table 7 – Comparison between different flowmeter elements

Flowmeter Element Recommended Service Range Pressure Loss

Typical Accuracy (%)

Relative Cost

Mass (Coriolis) Clean, dirty viscous liquids; some slurries 10 to 1 Low ±0.4 of rate High Mass (Thermal) 10 to 1 Low ±1 of full scale High Turbine Clean, viscous liquids 20 to 1 High ±0.25 of rate High Vortex Clean, dirty liquids 10 to 1 Medium ±1 of rate High Electromagnetic Clean, dirty, viscous conductive liquids and slurries 40 to 1 None ±0.5 of rate High Ultrasonic(Doppler) Dirty, viscous liquids and slurries 10 to 1 None ±5 of full scale High Venturi tube Clean, dirty and viscous liquids; some slurries 4 to 1 Low ±1 of full scale Medium Flow nozzle Clean and dirty liquids 4 to 1 Medium ±1 to ±2 of full scale Medium Target meter Clean, dirty viscous liquids; some slurries 10 to 1 Medium ±1 to ±5 of full scale Medium Positive Displacement Clean, viscous liquids 10 to 1 High ±0.5 of rate Medium Pitot tube Clean liquids 3 to 1 Very low ±3 to ±5 of full scale Low

Initially, two flowmeters will be required for the air and water lines before the arrival of fluids at the independent

manifolds. Regardless of economic issues, we decided to install a coriolis meter for the air line because of its high accuracy and low maintenance; as for the water line, we chose a turbine meter. With a coriolis meter, the estimated cost will be USD 7,000 with pressure drops varying from 0.043 to 13.2 psi, the turbine meter will cost around USD 2,000 with an expected pressure drop of 2 psi. Conclusions A new dedicated facility to be located at Texas A&M University will house experimental investigations of liquid loading in gas wells. A technical and economic analysis of the design options considered for the compression and pumping system showed that using a compressor and a water pump is the most feasible option for the Tower Lab.

For the compressed air system, a technical and economic evaluation identified the most attractive candidate as one with three-reciprocating units working in parallel. Each machine provides 580 psig and 168 scf/m and they are conveniently balanced with minimal vibration and low structure and air-borne noise.

16 SPE 122786-PP

The pressure vessel located at the bottom of the vertical section of the tubing was designed not only to minimize costs and weight, but also to facilitate its installation in the laboratory. The vessel will have a top and bottom flat head cover with multiple openings welded to the cylindrical shell body. These openings will allow access to the internal space in the vessel to accommodate glass beads representing the porous medium. The vessel also incorporates eight openings to attach the injection manifolds, fittings to install pressure gauges, and lugs for lifting and installation purposes.

The required superficial velocities for air and water showed the need for independent manifolds to be connected to the pressure vessel. Two options are recommended for the geometry of the device, but costs and ease for fabrication need to be addressed. We chose 3 in. and 1.5 in. schedule 80 for water and air, respectively.

The air and water pipelines were selected based on the superficial velocities required at the entry points at the pressure vessel. They will accommodate different types of fittings, such as long radius elbows and reduction/expansion joints, as well as pressure gauges and flow metering devices. The total pressure drop in the system was estimated to be below 6 psi for both lines including their respective manifolds.

Digital pressure gauges with a minimum range in accuracy from 0.05 to 0.1% are to be installed on the fittings located at the top of the pressure vessel, as well as the eight entry points in the injection manifold. In addition, two flowmeters will measure the flowrate of air and water. The selected metering technologies are coriolis for the air line and turbine for the water line. Acknowledgments We would like to express our sincere gratitude to the Crisman Institute at the Harold Vance Department of Petroleum Engineering at Texas A&M University for funding this project. Nomenclature

d = Pipe internal diameter, in f = Moody friction factor, dimensionless g = Gravitation constant, ft/s2

HL = Friction head loss, ft L = Length of pipe, ft p = Internal pipe pressure, psi p1 = Upstream internal pipe pressure, psia Δ p = Pressure drop, psi (liquid) psia (gas)

Qg = Gas flow rate, MMscf/d Ql = Liquid flow rate, bpd S = Allowable stress for pipe material, psi T = Flowing temperature, oR V = Velocity, ft/s Z = Compressibility factor for gas, dimensionless γ = Specific gravity, dimensionless

References

1. Alshuraiqui, H., Grattoni, C., and Muggeridge. A: 2003 Laboratory and Numerical Studies of First Contact Miscible WAG Displacement: The Effects of WAG Ratio and Flowrate. Paper presented at the 12th European Symposium on Improved Oil Recovery, Kazan, Russia, 8-10 September.

2. American Society of Mechanical Engineers, Boiler and Pressure Vessel Code, ASME, Section VIII. 2007: Rules for Construction of Pressure Vessels, Division 1, New York. 15-364.

3. American Society of Mechanical Engineers. 2004. Chemical Plant and Petroleum Refinery Piping. New York. 4. Coleman, S., Clay, H., McCurdy, D., and Norris, L. III. 1991: Understanding Gas-Well Load-Up Behavior. Paper SPE 20281,

JPT. 5. Coleman, S. and McCurdy D.: “A New Look at Predicting Gas-Well Load-Up,” paper SPE 20280, JPT March 1991. 6. Costantini, A. 2005: Dynamic Interaction Between the Resevoir and the Well During Well Testing, Dip. Ing. thesis, University La

Sapienza of Rome and Imperial College. 7. Corteville, J., Grouvel, M., and Lagiere, M. 1983: Experimentation des Ecoulements Diphasiques en Conduuites Petrolieres:

Boucles d’essais de Boussens,” Revue de l’Institute Francaise du Petrole. 38 : 143-151. 8. Dousi, N., Veeken, C.A.M., and Currie P.K. 2005: Modeling the Gas Well Liquid Loading Process. Paper SPE 95282-MS

presented at the Offshore Europe Conference, Aberdeen, 6-9 September. 9. Falcone, G., Teodoriu, C., Reinicke, K., and Bello, O. 2007: Multiphase Flow Modeling Based on Experimental Testing: A

Comprehensive Overview of Research Facilities Worldwide and the Need for Future Developments. Paper SPE 110116 presented at the SPE Annual Technical Conference and Exhibition, Anaheim, California, 11-14, November.

10. Falcone, G., Hewitt, G., and Richardson, L. 2003: ANUMET: A Novel Wet Gas Flowmeter. Paper SPE 84504-MS presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, 5-8 October.

11. Gool, F., and Currie, P. 2007: An Improved Model for the Liquid Loading Process in Gas Wells. Paper SPE 106699-MS presented at the Production Operations Symposium. Oklahoma City, Oklahoma, March 31 – April 3.

12. Guo, B., Ghalambor, A., and Xu, C. 2005: A Systematic Approach to Predicting Liquid Loading in Gas Wells. Paper SPE 94081-MS presented at the SPE Production Operations Symposium. Oklahoma City, Oklahoma, 17-19.

SPE 122786-PP 17

13. Hart, R. 2007. http://www.swri.org/4org/d18/mechflu/fluiddyn/pdfs/flowloop.pdf. November 2007. 14. Kegel, T., and Kinney, J.: “Wet Gas Metering Facility at CEESI,” Nunn, Colorado, Internal Report. 15. Langsholt, M. 2007. http://www.ife.no/laboratories/well_flow_loop/index_html-en/view. November 2007. 16. Lea, J.F., Nickens, H.V., and Wells, M.: Gas Well De-Liquefaction, first edition, Elsevier Press, Cambridge, MA (2003). 17. Neves, T. and Brimhall R.: “Elimination of Liquid Loading in Low-Productivity Gas Wells,” paper SPE 18833 presented at the

SPE Production Operations Symposium held in Oklahoma City, Oklahoma, 13-14 March 1989. 18. Robole, B., Kvandal, H. And Chuller, R.: “The Norsk Hydro Multiphase Flow Loop. A high pressure flow loop for real three-

phase hydrocarbon systems,” Flow Measurement and Instrumentation 17 (2006) 163-170. 19. Saaty, T.: “A Scaling Method for Priorities in Hierarchical Structures,” Journal of Mathematical Psychology, 15, 234-281 (1977). 20. Sintef. 2007. http://www.sintef.no/Home/Petroleum-and-Energy/SINTEF-Petroleum-Research/Wellstream-

Technology/Laboratories/. November 2007. 21. Snow, D.: Plant Engineer’s Reference Book, Elsevier, Second Edition, Burlington, Massachusetts (2003), 26. 22. Solomon, F. and Fernandez, J.: “Design and Construction of a Dedicated Facility for the Experimental Study of Liquid Loading in

Gas Wells,” Interim report for the Crisman Institute, College Station, Texas, 5, November 2007. 23. Stephenson, G., Rouen, R., and Rosenzweig, M.: “Gas-Well Dewatering: A Coordinated Approach,” paper SPE 58984 presented

at the International Petroleum Conference and Exhibition held in Villahermosa, Mexico, 1-3 February 2000. 24. Toma, P., Vargas, E. and Kuru, E.: “Predicting Slug to Annular Flow Pattern Transitions for Reducing Risks of Gas Lift

Instabilities and Effective Gas/Liquid Transport from Low Pressure Reservoirs,” paper SPE 100615 presented at the Gas Technology Symposium held in Calgary, Alberta, Canada 15-17 May 2006.

25. Turner, R., Hubbard, M. and Dukler A.: “Analysis and Prediction of Minimum Flow Rate for the Continuous Removal of Liquids from Gas Wells,” paper 2198, JPT, November 1969.

26. TUV NEL. 2008. http://www.tuvnel.com/content/pdfs/wet_gas_facility.pdf. November 2007. 27. Unander, T. E. Sintef. 2007. http://www.sintef.no/Home/Petroleum-and-Energy/SINTEF-Petroleum-Research/Wellstream-

Technology/Laboratories/ 28. Vilagines, R. and Hall, A. “A comparative Behavior of Multiphase Flowmeter Test Facilities,” Oil and Gas Science and

Technology – Rev. IFP, (2003) 58, No.6, 647-657. 29. Yamamoto, H. and Christiansen R.: “Enhancing Liquid Lift from Low Pressure Gas Reservoirs,” paper SPE 55625 presented at

the Rocky Mountain Regional Meeting held in Gillette, Wyoming 15-18 May 1999.

�������� ��� ������� �� ������� �������������������� ���� ��� �� �����

��������� �� ������� �� � ��������� ��� ������� �������

���� ��� �� �� � �� ��� �� ����� ���������� ��������� ��� �� ������ �� ��������� ���������� ������� ������ ��� ����� � ����������� ����� ��� ������� ��� � ����� ����� ������ �������� �� ��������� � ����������� ����� ��� ��� ��� �� �������� ��������� �� ��������� �� ���� ��� �� ��� ����������� ����� ��� ���������� � ����� ����� ������ �������� �� �� ������ ���� � ���� ���� ��� �� ���� �������� ����� �������� �������� �� ������� ����������� ����� ��� ���� ��� �� � ������� ������� ���� ���� ������� ��� ������������ ���� �������� �� �� ������� ������� �����!��� �� ������� �������� �� ��������� � ������� �������� ��� �� ��� ��� ��!������������ ��� ���� ������� ������ � �� ����������� � �� ���� ���� ������� ��� �� �� ������� ������������������� ��� ���� ��� � �� "# �������� �� �� ������� ������ ��� ������� ������� ���� � �� ������� ���������� ������������ � �� ����������� � �� ���� ���� ��� �� � ������� ����������� ����� ��� ���� ���� ��� ���������� �� ������� �������������� ����� ��� ���� ��� ��������� ��� ��������� � ������� �� ���������� � ��� �� ���������� ���� ���� ��� ����������� ��� �� �����!���� ������ � ���� �� ����� ������ ����� ��� ������� ���������� �������� ��������

� ����������

$�����!�������� ����� ��� ������� ��� � ����� ������ ������ �� ����� � ��� ���������� ������� � �������� �� ��� ��� �������� �������� �� ������������ ��������� ��� �������� �� ��� ����� ��� �������� ������ %&'()�*��� ���� ��� �� �� � �� ��� �� ����� ����������� �������������� ��� �� ������ �� �� ������������������� ������� ��� �������� �������� � �� ���������������� ��� ��� ��� � ����� ����� ������ �������� ������������ �� �� �� ���������� ��� ������� � ����

��!�������� ����� ��� ���� ��� � ������ �������� �� ������������� +�� � �� �� ���� ������������� ��� ���������� ������ � ����������� � ������� �� � ��� ��������� �� ����� ��� �� ���� ������ � ��� ������� %,'&()��� ���� � �������� ��� �� � %&,'-.) �������� ��������� ������� � ���� ������� ������ �� ������!������������ ��� ���� �� �� ������� ������ �� ������������ ������������� ��� ���� ������� ��� ���� ������� � ��� ���� ��� ���� ������ ��� ���� �� ��� ������������ ���� ����� ��� ��������� ���� ������� �������� �� ������!��������� ������������ ��� ���������� ����� �� �������������� ���� ������� ��������� ���� �������� ����� � ������ ������� �������� �����!��� "� ������ ����� ���� ������ �� � ������ ���� ������� ��������� �� �� ���� ��� ����������� ����� �� ����� ���� � ���� ���� ����� ������� ����� ���� � �# ��������� ��� �������� �������� �����/������� � ��������� �� ������� ���� ��� ���� ��� � ���� 0������ ���� ������������ ���� ������� ���

����� �� ���� ���� ��� �� � ������!�������� ����� ������ ���� ������ ������� �� ����������������� �� ��� %-1'-2) ������� �� �������� ������� �

������� � ��������� ������ ��� �� ������ � ��������� �� ����� �������� � �� � ��������� ������ �� ������� ��� ���� ������!�������� ����� ��� ���� ���� �������������� ��� �������� �#��� ����� ������ ��� ���� ������������� ��������� ����� � ���� ��� ���� ������� ��� ���������� ������ � �� ��������� ���� �� ����� ���������� �� ������� ������ � �� ���� ��� ��� �� ����������� ��� � ���� ��!��� ��� ���� ��� ��� �������� ��������� � �� !����������� ����� ����� 3�� ������ �������� � ��� �� ����� � �� ���� �������� ��� ����� ���� ���� ������� ��� �������� ���� ������� ��� ��� ���������� � �� � ���� � � �������� ������!�������� ����� ��������� �� ��� ��� �� ���� ��� ���� ������� ��� �� ��� �� ������� ���� ��� ��� ���� � ������ �� ����������� ����� ��� ��� ��� ��

����� ����� ������ ������� � � ������� ������ �� ���� ������ ��������� � ��� �� ���� ������ � �� ���� ��� ��� � ���� ����� ��� ������� ��������� 4�� ������������ ��� �� ���� ������� ������� ������������ ������ ���# ��� �� ��������� �� ��������� ������� �� ������ ��� � ������� �� ���� ���� � ���� �� ����� ������ ����� ��������� ��� �� ���� �� ������� �� ��� ������ ���������� � ����������� ����� ��� � ����� ����� ������������� ��� �� ���������� *��� ���� ��� �� ������!�������� ����� ��� ���� ��� ������� ������� � ����� ��� ������� ������� ��� �������� ������� �� �� ������ ���� �� �� �� ����������� �� ������� � ����������� ���������� ������� ��� �� ����������� � ���� ������� ������ �� � ������!�������� ����� ��� � ����� �������� �������� �� ���� �� �������� ������������ ���� �������� �����!�� � �� ���� ��� ������� ���� ������� ������������ �� �� ����� ��� � ����� ������� 5��

&1.6 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ 5=�> &8�&88-?�����-88188&21 ����� ���� �������� ����� ��� @� &-

'

��� 5� ������� =� =� ���� A�������B� ��������C� 5������� <� +� D��������5������� /� ������� ��������� � E������� "����������� ��������F��������G� /�������� ����������H� &8� 5�I,6(, /��������J����������$�������

��� �����

��� ���� ��� � ������� � ����������� ������� ������������!�� �� ���� � ��������� ���� ���� ��� ������������� �� ������� ����� ��� ���� ���� � ������ ����������� ���������� ������� ������ � ������ ��� �� ��� ����������� ����� ��� ������� ��� � ����� ����� �� ���������5������ ������� �����!�� �� �� ���������� ���� ��

����� ��� ��� ������������ ��� �������� %I8'I6)� �� �� ����������� ��� �� �#������ ����� � ���� ��� ��� ������������ ������� � ��� ����������� ��������� ������ ��� ��� � �� �����!�� K�������� ��� ������ � �������� ����� ��� ����� ������ 0����� ����� ���� � ������������� ��������� �����!��� ��� ��� ����� ����������� ���� �� ���� ��� ��� ��������� ��� ��� ������ � ���� � �� ��������� � ������� � ������������� ���������� � ��� � ���� ����� �� �������� ����� ��� ������� ��� � ����� ����� ������� �����# ��������� ������� ��� �� ���� �� � ����� �� �� �������� �� �������� ����� �#������ ����� ��� ���� ������ ������ �� ���� ��� ������ ��� ����� � ������� �� �� ����� ��� �� ��������� � ����� ��� ������� � ����� ����� ������ ������� ��� ��� ������ ����� �������� ������ ��� �������� �� ���� ����=�� ��� ��� ��#����� ��� ������� ����� ��� ���� ������

���� ��� ��������� � � �������� ������� �� ��������� ��� � ������� ������� ��� ������� �#��� �������������������� ���� ��� ������� ��� �� ��� �� 3��� &� +����������� ����� ��� � ������ ������� ��� ��� ���������� ����� �� �� ���� ��� ���� ����� ��������������� ��������� ���� ��� ��������� 0�� ���������� �����#��� ����# �� ���� ��� ��� ���� �#��������� �� ������ �� ���������� %I('I,)� �� ��� � �� ������ ��� ���� �� �� ������� ������� �����!�� �� �� ����������� ��� ���� ���� ��� �� � ������!�������� ����� ��� �������� �� ������� � �� ��� ��� ��!��� �� �������� ������������ ���� ������� ������ � �� ���� ���� ��� ���������� ��� ������������ ��� ����������

� �� ��� �����

�!� " #���$�

$�����!�������� ����� ��� ��� ����� ���� ������� ��� .8 �� �������� �������� A�5C ��#������ � � ��� ������������ ��� ���� ������ ��� ����� �� �������� A��� 3��� -C ��������� � � 6�8 � ��� ���������� ����� �� �� � � ������ �� ������ ���������� ������� ������� ��� ��������������� �� ����� � � ������� �� ��� �� ���� ������ ��� ����� ����� ��#������ � � � .8 �� �������� �������� ���6�1 � ������ �� � � ��� ���� � ��������� �������� - ����� �������������� ������� ������ � �� ���� �� �������� �� ����+����������� ���� ������� �� � �������� � 1 � ���

�� ���� ������ ������ � ���� �������� ������� ��� � ������ ����������� ���� � �� ��������� ������ �� � � ��� ���������� �� � ����������� �!�� � �� �# �� ���� � ������� ����� ��������� 4����� ������� ���� ��� ���� ��� ������ ��!���� ��� ��� ���� ����� ��� ��������� �� ����� ����� ���� �� � 6�8 <4 ��������� ����� ����������� ������ �� ��� � 8�. �I ��������� ����� ������ ���� ��� � -1 �� �5 ���������� ����� � �� ���� ��� ��� ���� ��� �������� ����� �� ��������������� �������� ��� ������ �� ��� ���� � �� �������� �������� �� �� �������� � ��� � �� ����������� �� � �� ��� �� �� ����� ��� ��� � �� ��#��� ������� ���� �� ����� ������� �� ���������� ������ ��� �� ���� � �� �� ��� ���� � �� ��� ��� �������� �� �� ����������� ��� �������� �� � ���������� �������� ���� ��� ���� �� ����� ���� �������� �� ���� ��

"��� 3������� 0����� $��0� $������� ��� � ��������� �������� � 8�6 �� ��� � ������� � -688 ��?�I� ������ �������� �� �� �� ��� �� ���� � �� ����� ������ �� �������������� ������ ��� ��� ��� ��� ����� ����� ���� ������ �� �� ����������� �� � �� ������� ������� �� ���� ��� �������� �� �� �������� ���!����� � ���������������� ������� �� ���� ������� ���� ������������ ��� ����� �� � ��� ��� ��� ������� ����� ������ ��������!�������� ���� ��� �# �������� � ��������� �� ������������ � �������� ��� ����� �� �� ������ ��� ��� ������ ���� ���� �������� ���� �������� �� ������������ �������������� ��#���� ��� � ���������� ��������� �������� � �������� ������ 4�� �� �������� ����� ��� ����� �� �������������� ��#���� ��� ��������� ��� ����������� ����� E������ ����� � �������� � ������ %I2) ���������� ��� � ��� �� �������� ��� ��� ���� ��������������� ������ ��� �������� �� ����� ���� ������ �� ������������� �������� �������� ���� �������� �� � ������� ����������� �� ��������� ��� �������� ���� ����� ����� � � ���������� �� �������� �� ��������� ����� ��� �������� �� ���������� �������� ��� �� ���� ������� ������ ��� ������������ �� ������������� � �� ���� ������� ������ ������������ ��� ����� � 8�88- ��?�I��� ���� ����� ��� ���� ��#���� ��������� ��� ������ �

��� ��� �� 8�8. � L 6�1 � ����� ����� ��#������ ���� � ���� ��#���� ���� ���� ��� �� ������ ���� ��� �� ��� ������

����� ���� �������� ����� ��� @� &- �� >??��������K�������� 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ &1.(

��$��� �! �� ���� ��� ������ �� �� ������!��� ��� ��� ��� �� �������� � �� %I6)�

��� �����

� �� ���� ���� �� ���� �������� ���� �� ������ ����� ����� �� ����� � � ���� ������ ������ ��� �� ����� ������������� ��� �� ������� @��� ������� �� � ���� ��������� ��� ���� � � � ������� �� ������ ��� ������� ������� ������������� ���� �� ���� ������� ���� & ����� �� ����� � ���# ��������� �������� ������������ �� ��� ������

�!� #�$���� � �$� ������� ��� ��� %����������� &'���

������ � �� ������!�������� ���� ��� ���� �������� �� ����� ����� � :5� :���M��� ���� ��� /+=�0//�&888 ������ ��� � ���� �#�� �������� ��� ���������� ���� ��� ���� ��������� �� :5� :���M��� ������������� � � ���� ��� ������� ���� ���� ������ ��� 0//������ ����� ��� ����� ������� ���� ���������� ��� ����� ��� �������� �������� � ����� �������� � &8-. L 1&-��� ����� ��� ����� �� �������� � ������!�������� ������� � �� ������ ��� � ����� �������� ���� ��������� ������ ������� ��� ��������� ��� ��� �������������� ���� ����� ��� ���� ���� �� ���� � � ��� ����������� ��� &888 4��� ���������� ��� � �� �� � ��� ������ ��������� ��� ��#�� �� � ���� ����� � �� ���� �� ������

&1., 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ �� >??��������K�������� ����� ���� �������� ����� ��� @� &-

��$��� �! �������� ������� � �� �# ��������� ���� � "+> "������������� ������N 3�> 3��� ������ ����N��3> ���� ������ ������N 3+&> 3�� ����� A���C N 3+-> 3�� ����� A��!���CN ��> ���������� ����N :&> ����������� &N :-> ������ ����� -N :I> $�� ����� &N :.> $�� ����� -N :1> $�� ����� IN +> $��������� ��#��N E�>E�� N /> /� �����N :6> 5������� �����N 3> �� ������N ��> ���� ������N E�> E������� �� �N 5E> 5������������ ������� ����������N 5> �5 /�������N 5�> 5��� ��!������� ������N 4*&> 4��� ���� �����������N4*-> 4��� ���� �����������N 0�/> 0���� ��� ������N 0�:> 0���� ��� ���� ������N 3$> 3�����������N E/&> E������ �� ���� �� ������� ��!�������N E/-> E������ �� ���� �� ������������ ����������������� ��!��������

(�)�� �! "# ��������� �������� �� ���� ������ �� �# ���������

� � �������� A�C 8�8.

� � ����� A�C 6�18

� � ����������� �������

������� ���� ����� A��C 8�68'8�,8

������� ���� �������� A��C 8�68

������� ������� A��?�IC -688

������� ������ A��?�IC 8�&8'8�,8

�������� ������� �������� ������� A�?�C 8�82

������� ������� �������� ������� A�?�C 8�&.

������� D������ ������ ,.�88

������� ��� ��� ���� A������������C -I�88

����� ������� A��?�IC 22,

����� ��������� �������� A& O &8'6 �-?�C &�8-

�� �������� ��� ������� A�?�C 8�.8'&�68

�� �������� ��!��� ������� A�?�C 8�18'&�18

��� ��������� �������� A& O &8'6 �-?�C &1�&8

����� ������� ������ A@?�C 8�8(8

��� �����

�� ������ ����� ������ &-, ������ � �� ������� ����� ������ �� ����� ���� ������� �� �� ���� ����� ����� ���� ������ ��� �� /+=� ������ ��� ���� � � &(����� �� �������� �� ���� ������ ��� ������ ������� ������� � �� ����� !������ ��� ����������� �� ������ �������� ��� ������� � ������� ���� � �������������� ����� ��� � �� ���� �� ������������ ����� ��������� ���� ���� ��������� � ��� �� � � ���������3� ���������� �� ��� � ��� � �� ��� ��� �������

��� ��� ���� ����� ��� ������ �� �#�� ��������� �������� ��� �����!���� ����� �������� � �� �������� ������������� �� ��� ������� ��� ������ ��������� A�������������C �� �� ���� ���� �� ������ � ��������� ������� �������� ���� ������ � ������������ ����� � � �#�������� I(� ��� ������ ���� ������� ������� �� �������� ����� ��������� ��������� � ���� �������� ������� ������ ����������� ������� �� �������� � �� ���� ������� �������� ������ ������ ���� ��� ����� �� ���� -�

�!* #�$���� � �$� +���'���

�� �������� � �� ���� ������� ��� ��������� ��� ����� �������������� �� �� �������������� ����� ��� ����

��� ��!����� !����������� ���� ���� ������� ��� ������������ ����� "#������� � ���� ���� ��� �� ������� ��������� ��� �������� �� ��� � �� ������ ����� �� ��������� ������� ������� ������� ������ �� :5� :����M��� ������� ��� ���� �� ��� ������

������� ��� � &��� ��� ������

@ ��������� ��������� �� ���� ��� ������������� ����� ������!�������� ���� ��� �� ��������� ����������0����� � �������� ������� ��� ����� ����� �� ���������� � �� � � ������ �� ���� ���� ��� �������������� � ���� ������� � ���������� ������� �� ���� ���� ������� ��� �� �� ������� �������������� ����� ������� ���� � ������������� ��� ��� �������� �� ����� 3��� I ��� �� ���� ��� � �� ����� ��� ���� �������������� ��� � �� ���� �������� ��� � �#�������

� ������� �� ������ � ���� �������� �� ��� ������ �������������� ����� �� ������ �� �� ����� � � � ��� ������ ���� ��� �� ������ ������� � ���� �������� �� ��������� � �� � �� @�#�� �� ����� � ��� ������ � ������ ��� ����������� �� ���� ��� � ��� ������ �

����� ���� �������� ����� ��� @� &- �� >??��������K�������� 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ &1.2

(�)�� �! 3������� � �� :5� :���M��� ���� ��� /+=� 0//�&888 ������ �����

����� �������� � � &8-. L &8-.

����� ���� 2-I � �

������ � ������ &-, ������

�#�� ���� &8 O &8'6 �L &8 O &8'6 �

������ ����� ���� -�16 A0C ��L-�16 A:C ��

����� ����� ������ /+=�������

����� 1&- +�����

������ ����� ��������� ������ ��� �# ���� ���� ��� - ��

������� �#�� ��������� � �� ������� ����� ��� ������ ��� ����� � �� ������

������� �� �� ,���� A��C

�� ���� ����� D��-I- ���������

�� �� ����� ���� � � .888 ������ �� �����

�� �� �#�� ���� II +0�

��� �� �� &- : 5/� 188 �

������� ����� ���� ���������� �������� �������

���� ���� �������� /�����

������ ����� ������� �����# ����� ����� �������

������ �������� ���� ����� �� � K ��� ���� ��� ���

�������� �������� +������ �#��� �� ������ �������� ������� ������ ��� ������� ������������������� ����� ����� �� ���� ��� ����� �� �������

��� �����

�� � � �� ���� � ��� �� ���� ��� �� ���������� �������>&C

�� �����

A&C

�� ��6� �I�� A-C

�� P �!" AIC

" �#-

.� �� ���& - � ��� &� � � �- � ��� &� �

�������������������������������

&� �- � ��� &�-�

� �

A.C

�� ���� ���� � ������ & �� ���� �$ ��� �� ��������������� � ������� & ��� - A���� ��C ��� �� ���������� �����"!� A.C� �� ���� � ������ - �� ��� ������� �� ������������� ���� � ������ & ��� �� ���������� ����� �� ������ ��� �� ��# ������� � �� ���� ��� �� ������� �� ���������� �� ����� � ��� ������ �� ��� ����� �� �� ����� � ��� ���� ��� �� ����� � �� ������ <����� ������� � ��� ������ � �� ���� ��� �� ���� ������� ������ �� ��� ������� �� ���� ��� ����� ��� �� �����������> �� ������� ���� ��� �� ��� ��� �� �� ���� ��� ���� ���� ����� ��� �� ���� ����� � �� ����������������

* ,������ ��� #��������

*!� %����� -)���.�����

���� ������� �������������� �� �� � ����� ��� ������������� �� ������ ������������ � �� ���� ��� ������ �� ��

�������������� ����� ��� ���� ��� ������� � �#�� �� ��� ������ ���� �� �� �������� �� ����� �� 3��� .� ������ �� � ����������� ��� ���� ��������� �� �� ��� ������������� � ���� �������� ����� �� � ������!��� ���� ������� � � �> � ���� ������� ����������� ��� ��� ����������� ���� ��������� � � ���� ��������� ��� ���������� �������� ��� �� � ������ ������� �������� � �� ���������������� �������� ��� �� ��!��� ��� ��������� ��� �������� �������� ��������� �� �� ���� � �� ��� ���������� �� ���� �����

�� ����� ��� ����� ��������� ���� ��������� �� ��������� ����� �� �� ���� ������� � �� ���� ������������ ������� ��� ��� ������� � ��� ��������� �� ������������ ��� �� ��������� �� � ������� � ���� �������� ���� �� ��#��� ����� ������� � ��� ������� ��� �� ����#���� � �� ����� �� ������� � ��� ������� ��� ��� �������� ��� ���� �������� ��� ������� ���������� ��� ������������� �� ���K����� � �� ���� �������� � �� ���� �������� �� ������ � ��� ������� ��� ���� ������ ���#��� ��� � �� � � �� ����� ������� � �� ������ � �������� ������ ��� ����� � ������ �� �������� � �� �������� ����� ��� ������� �� �������� ��������� � �� ��� ��������� �������� ������ ���� �� �#�� � �� � �� ��� �� � ������������� ������ ������� ���� ������ ������������� %-1� I&)� � �� ���� ��� � �� ����� ������ � �� ��!��� ���� ����

������ ��#��� ��� ������ ���� ������� ������� � � ���������� � �������� ���� ������� ������ � �� ��� ���������� ������ �K����� � ���������� ����� ���� ����� ���� � �� ���� � �� ����# ��� ������ � ����������� ����� ������ �� ���� � �� ��� ����� �� ���

&118 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ �� >??��������K�������� ����� ���� �������� ����� ��� @� &-

Flow

direction

��$��� *! ������������� ���� � �� 6������� ��� ����� ��� �� � ��� ����� ���������

(a)

(b)

(c)

(d)

(e)

(f)

��$��� /! ������ ����� ���� ������� ����������� �������������� �� ������������ ��� ���� ��� �� � P 8�68 ��� %� P 8�6 �?�� ��� %� P &�8 �?� A�� P 8�8I�8�81� 8�8(� 8�82� 8�-� ��� 8�I1 ��?�I �� ����� A�C � A�C� ��� ��������C�

'

&C *��� � ������ �� �� ��� � �� � ���

��� �����

����� ����� ����� �� �#�� � �� ���� �������� ��� ��������� � ��� ������� ����� �� ���� �������� ��� �������� �� ��� ������ ��� ������� ���������� �� ���� ������������������ �� �� �������������� ���� ����

*!� "��� ������������ &��� ��� �������

�� ����������� ������ � �� ������� ����� �������� ������!�� ������ �������� �������� � ���� ��� �������� �� �������������� ����� ��� ���� ���� ����� ����������� �� ������� �� ������������ ���� ���� ��� ���� �������������� ����� ��� ���� ��� �� ��������� �� ��������� ��� ��������� ��� �� �� ������������ #��� ������� �� ������� ��� ��� � �� ���� ��� �� ����������� �������� ��� ��������� �� � ������� �� �������� ��!��� �������� ��� ���� ������ ������ �� ��� �� 3��� 1� 3� ��8�181 ��� 8�686 ��'& ��� ���������� �� ������ ��� ���� ����� ���� ��� ����� ����� �� �� ����� ��� ��� � ��� ���� ������� ���� �� ����� ������� � 3��� 1 �� �� �������������� �� �� ���� ��� ����� ������� ����������� ���� ?5 8�. ������� �� �� ���� ��� �� ���� ��� �������������� �� � ��������� ������ ����� �� ������� � �� � �� �� ������� ������� ���������� ���� ����>

� 5� ������ � �� �� �������� ��� ���������� �� ������������� �������� �� !���� �����N �� ���������� �� ������ ���� ��� ��� �� ���� -1Q�

� �� ������ � ����� ������������� ��� ��� ��� �������� ���� ���� ��� �������� �� ��������� ��������� ��������� � �� ���� ��� �� �� �������������� ����� ��� ���� ����

3��� 6 ��� ��� �� ������� ��� ��� � �� ���� ��� � ������ �� �� �������� ��!��� �������� ���� �� �������� ��!���� ��������� ���� �� ���� �� ��������� ������� �� ��������� � ���� ��� �� �������� �� �� ������ �� ��� ���� �������� ��!��� ��������� ������ ������������� 3��� 6 ������ ��� �� ������� �� �������� ��!��� ������� ���� �������� ���� ���� ��� �3��� ( ��� �� �� ������� � �� ���� ��� ��� ��

���� ������� � ���� ������� � 8�I ��� 8�. ��?�I� �� ����

��� �� �� �������������� ����� ��� ���� ��� ��� ������� �� ���� � �� ���� ������ �� 8�&8� 3��� ( ������������� �� ��# ���������� ������� �� ���� ������ ��� ������ ���������� �� �������� �� ��������� � �� �������� ��� �� ������ �� �� ��� ������� ������� �� ������� �� ��!��� ��� � �� ����� ��� ��#����� ��� ���������� ���� � �������� ������� �������� ��� ����������� ���� ���� �# ����������� �� ������� ����������%-6� -,� -2)� �� �� ������� � ������ ���� ����� ��������������� �� ���� ��� ������ ������ �� � ����� ���� ������ ��������� � �� ���� � �� ������ �����!����� �������� � �� ����������� ������ � �� ������� ������

*!* "��� ���� &��� ��� �������

3��� , ������� �� ���� ��������� ���� ��� �� � �������� � �� �������� ��� ������� �� ��������� �#��� ��������� ����� ����� �� ���� ��� ����� ��������� ��� ����������� �� �������� ��� ������� ����� �� ������ ������������ � �� � �� �� ���� ��� ����� �!���� �� ���������� ������� � �� ���� ��� ���� �� �������� �� ���%-1)� �������� �� ��� %-,)� ��� +������� ��� �������� %I8)� �� ���� �������� ������ ���� ��� ��������� ���������� �� ����� ������� ��� �� ���������� � �� ������ ��� ���������� ��� ����� ������ ����� ��� �� ���� ����

����� ���� �������� ����� ��� @� &- �� >??��������K�������� 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ &11&

0

0,000005

0,00001

0,000015

0,00002

0,000025

0,00003

0,000035

0,00004

0,000045

0,00005

0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0

h/D

so

lid

pa

rtic

le h

old

up

jg = 0.505 m/s

jg = 0.606 m/s

jg = 0.710 m/s

jL = 1.1 m/s

εs = 0.1 kg/m³

Jg

Jg

Jg

��$��� �! "����� � �� �������� ��� ������� � �#��� ����������� � ���� ��� �

0

0,000005

0,00001

0,000015

0,00002

0,000025

0,00003

0,000035

0,00004

0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0

h/D

So

lid

pa

rtic

le h

old

up

jL = 1.10 m/s

jL = 1.16 m/s

jL = 1.21 m/s

Jg = 0.505 m/s

εs = 0.200 kg/m³

JL

JL

JL

��$��� 0! "����� � �� �������� ��!��� ������� � �#��� ����������� � ������� �

0

0,000005

0,00001

0,000015

0,00002

0,000025

0,00003

0,000035

0,00004

0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0

h/D

So

lid

part

icle

ho

ldu

p

es = 0.10 kg/m³

es = 0.30 kg/m³

es = 0.40 kg/m³

jL = 1.100 m/s

jg = 0.505 m/s

εs1

εs2

εs3

��$��� 1! "����� � ���� ������ � �#��� ����������� � ���� ��� �

��� �����

������ ��� �� ������� ��#���� �� ������ ���� �� �������������� �� � ��K� ����� �� ��� ������ �� ����� ��#���� �� ������ ������������ %.8'.&)� �� ���� ���� ����� ����������� ��� ������� �� ���� �� ��� ��������� � ������ ��� ���� ������� ������� �� ����� � ���� ���������� �� ��� � �� �� ������� ����� � �� � � ����� � ��������� �� �#��� �������� � �� ���� ��� ������� �� �������� ���� ��� � �� ����� � �� 3���� 1', � ���� � �������� ��� ��

��������� ���� ��� ���������� ��� ������������� ��������� ����� ���� ��� ��������� ������� �� �� ���� ���������������������������� �� ��� ���� ��� �� ��� ������������ ��������� � ����������� ����� ���� � �� �������������� ������� �� ��� ������ ��������� ��� ��� ������� �� ��������������� ��� ������ ��� ��������� �� ������� ���� �������� ������� � ������� ��� 3�������� %.-)� ������� ��� �� ���� ��� ��!��� ������� ������������ �� ��������!�������� ����� ��� � ����� ���� �� ���������������� ��� ������������ ��������� ��� �� ������������� ����������

/ 2��������

�� ������ � �� ���� �������� �� � ������� �������������� ���� ��� ��� ������� ����� ������� ������� ������!�� �� ������ ��� �������� � ����������� ����� ��� ������� �� � ������ �������� � ���� ���������� ����� A��� ��������� �� ���� ������C� 5������ ����� �������������!�� ���� � �� � ������ �� �� ����������� ��������� ��� ����������� �� ���� ���� ��� ���������� �������� ������� � ����������� ����� ��� ������� �� � ������� � �!����� �� ���� ����� �� � ��������� �� ���� ���� ��� � ���� �� ��� ��� ��� � ����� ������������� ���������

���� ���� ������������� �� ���� ��� �� � ���� � ���������� ������� � ���� ��� ������� �� ����� ��� ������� ���������� ����� �� ���� ��!������� %.I) ��� � ����������� ������������ ������� %..)��� ������� ������� ��� �� �������� ��������������

����� ��� ���� ��� ������ �� �� ��������� .8 �� � ������ ��� ���������� �� �����>

� �� �#��� ����������� � �� ���� ���� ��� ��������� �� �#������� � ����� �������������� �� ��������� �� �� � � ����������� ����� �� ������ ������� �������� ��� �� ���� ������� ��������� �������� ��� ��� ��!��� ����������

� �� ����� ������ ��� �� � ��������� �� �� �� ��������� ��� ��� ��!��� ��������� �� ���� �� �� ���� ����������������

� ������� � �� ������� ��� ���� ��� ����������� ������ ��������� ��� ��������� ��� ��� ������ ��#������ � ��� ������ ����� ��� ��!��� ���� ����� ����� ����� ��������� �� �� ���� ������� ������������

� �� ���� ���� ���� ��� ����� ������ ��� ��� ������� � ����� ������� �� �� �������������� ����� ��� ���� ����

� �� ��������� � �� ��������� ��� ��� ������ ��#��� ��� ���� ��� � ���� � �� ������

+��� ��� �# �������� ��� � �� �������� � ����������� �� ���� �������� �� ����������� ����� ��� � ������ ������ ��� ����������� ��� �������� �� �� ���������� �������� �� ����������� 4�� �� �� ������ � �������� ��������� ������ � ���� ��� ������ �� ������!������������� ��� ���� ��� ��� � �� �� �� ������� ����� ��������� �����!�� ��� � � ����� �������������� �� ���� ����������� �� ������ � �� ������!��� ���� ��� ����������� ��������� � �� ���� ����� ���

+�3�����$� ����

�� ����� ���� ���� � ���� E������ D����� 4������� A�������� F��������� � /�������� $������C ������ �� ��� ���������� ��� �� ����� �������� �� �� ������ ���������� � +������ "� 4������ +� 5������� �� <����J� E�������� ��� E� E���� �� �� ���������� � �� �# ���������� �������� ��� ���� �������� �� ���������� � ��������� ������� ���� ���������� ����������� �� ��������� �� �� ��� $����� ������� "#����� ������� A5 5C ����� ������� �

D�������> R��� &-� -881 %/"� 8&21)

&' )�� ����

" %�) ���� � ��� ������ ����� ��� �� � � �� ����� ��������

# %�) � � ��������� %�) ������������� ������� ���������! %�) #��������� ����� � �� ������! %�) �#��� �������� ����� � �� ������� %�) ���� ������� ���������� %') ���� ������������ ���� ��� �� %') ���� ���� ���� ��� � %�) ���� � � � ��� ���� � �� � ��

%�) ������� ���� � �� ������� � �����

&11- 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ �� >??��������K�������� ����� ���� �������� ����� ��� @� &-

0,000000

0,000005

0,000010

0,000015

0,000020

0,000025

0,000030

0,000035

0,000040

0,000045

0,000050

0 0,2 0,4 0,6 0,8 1 1,2 1,4

Superficial gas velocities (m/s)

Lo

ca

l m

ea

n s

an

d h

old

up

h/D = 0.0 h/D = 0.2

h/D = 0.4 h/D = 0.6

h/D = 0.8 h/D = 1.0

��$��� 4! *��� ���� ���� ��� �� ��������� �� �������� ��� ��������� ������������ �#��� ������ �� �� ���� ����

��� �����

%& %�?�) �� �������� ��!��� �������%� %�?�) �� �������� ��� ������� %') ����� ������������ ����� %') ������ � � ������ ���� ��������

���� �� '��� %�) ���� �� ��������� ��� ����� ������ %�I) ����� � ���������� � �� ���'� %�I) ����� � � � ���'� %�I) ����� � � ������ �� �� ���

(���� �)�*���

�� %��?�I) ���� �������� %��?�I) ��!��� ��������� %��?�I) ���� ������� �������

,���������

%&) +� +� ������� �++����� ���������) ���+� �,--� 0����� �S� +���554�

%-) R� ������� +� �� 5��������� 3� ������� � 3� R� ����������� .�� "��������� ���+� ��� �!��*�� @�� =������� �� ��?=��� �����

%I) +� �� 5��������� �� "�������� �� ����� �--$ .��/#�� 0����'�� �������'��) .)��� ������ =< �����

%.) +� �� 5��������� R� �������� 3� ������� � 3� R� ����������� 0��� ���+� ����������� #�����1 .�� 2�3�4� * �����

%1) +� �� $��������� +� �� 5��������� %� ���� .��� ���� �551� $3� 1� A� ����������> @��� 4��� ��� 3������ 5����� ��� D���������C

%6) D� <� ������� +� �� 5��������� 3� R� ����������� R� ������� �� ������������� ��� ��*��� �������� .�� ���+�� E������ ��� �����

%() =� =� ����� �� =� 3������� �� ����� �+ ��� 55�� "���� ���+� ��� ������ �+ ������� .��� ����� ����� ������� @������� &2?-8 @�����*�

%,) +� 4����� �� "�'����� �� .����6&�7�� ���8 �� ����� ��� ��� "����������A"�> �� J����C� E������ E����� =#��� �51��

%2) �� ������� =� <�������� +� *������� �� =����� �� ����� 9�� 0��� ���+��� �������� ���8� �0D � @���� �545�

%&8) +� *������� �� @������ �++����� �������� ���+�� 0����� �S� F� �55��

%&&) /� � ��� D� $� $������� �� R� <������� +� 0� ������ �� ����� 9���������� ���+� �+ ��� .��� .��������8�� .������� �� E������������� � /�+� /������ D������ /������ =����� �55��

%&-) E� =������� ���+� .�� ��������� ��� ����������1 .�� �2$9�� @���55*�

%&I) D� $� $������� +� R� +�������� /� ��� %� ���� ���� �������� �55�� 59:,;� 16�

%&.) D� $� $������� +� R� +�������� /� ��� %� ���� ���� �������� �551� 54�I6�

%&1) D� $� $������� <� �� 0���� +� R� +�������� /� ��� ��8��� ���������555� $-9� -62�

%&6) R� D� �� ����� $� 0� E��������� �� ����� ��� 0��� ���+� �� �����������8� �� 0� D� $�� � /������ 3����� �551�

%&() +� R� R� <���� /� E� 3�������� �� R� 0���� %� �����) ������ �������� �����$�5� -88�

%&,) E� ��������� D� �� �� �� �� ����� ,�� 0��� ���+� �� �������� ���8��� 0� D� $�� � /������ 3������ R��� �555�

%&2) E� ��������� E������� ����� �� �� E� �� �� ��� ��� 3���� ���#�������� F��������� � /�������� �����

%-8) E� ��������� D� �� �� �� ��� (�� %� ����� $-- :5-;� .(�%-&) E� ��������� D� �� �� �� ���� %� ����� ���� ���*� �$� -(&�%--) E� ��������� R� "� <������� D� �� �� �� �� ����� $��� 0��� ���+� ��

�������� ��������� �� 0� D� $�� � ������� /����� �����%-I) E� ��������� D� �� �� �� R� "� <������� /� +�5������ ����� ����

.��� ����� 24� -&.2�%-.) E� ��������� D� �� �� �� R� "� <������� /� +�5������ �� ����� $-��

0��� ���+� �� �������� ���8� �� 0� D� $�� � /������ 3����� �����%-1) �� ��������� 0� +�������� � �������� �� ����� �+ ��� �!�� <���

�����+��1 ���� �������� ��� �������)������ A"��> +� $��� 3� +��������� $� E� /�����C� �������� ������� �551� � &&I('&&..�

%-6) �� ��������� 0� +�������� <� ������ 9� <��� @� <����� �� +�������� �� ����� �+ ��� %. �/". � ������� ���� %���� 0��� ���+��0������ 0����� �541�

%-() �� ��������� 0� +�������� �� ������ <� ������ �� ����� �+ ��� %����6=. .������ �� �8�6����� ���8 #)������� =���� R� �� �544�

%-,) �� ��������� 0� +�������� � �������� 0� �������� 0��� ����� %����� ���� �55�� �� I(�

%-2) �� ��������� 0� ��������� � �������� 0� +�������� 0� ���������� ����� �+ ��� $�� %. �/". � %���� 0��� ���+� �� ����� ����� ����R� ��� @������ �55��

%I8) 0� +�������� �� ��������� �� ����� �+ ��� 5�� 0��� ���+� �� �����������81 0� � >,�� *��� 3����� �554�

%I&) *� <��������� $� :��K���@�������� ����� ���� .��� &221� 2- :�-;�I-,1�

%I-) �� +��������� T� 5��� �� ��������� ���� %� �������� �������� ���/� 95:$�;� I8�

%II) 9� @��� +� 0� $������ %� ���� �������� �550� 5�4� -,1�%I.) 5� <������ $� 0�������� �� �����K��� ��)�� ����� �55�� 3 :2;� &821�%I1) 5� <������ $� 0�������� �� �����K��� ��)�� ����� �55�� 3 :2;� &&8(�%I6) 9� ������� � "� 5������ "0��� %� �510� �� :$;� .(�%I() +� $ ��� 4� E� R� ��� 0��� %� �������� ���8 �551� �5 :2;� 2.1�%I,) +� @������ 5� +����� 0��� %� �������� ���8 �55�� �$ :�;� -1I�%I2) :� +������� 3�� +������� � ���������� +�#����� �� E� �������

���#� ������� �������� F��������� � 5���� �551�%.8) +� /���������� 3� *������ +� @� +����� /� $��� R� /����� 0��� ����

����� ���� �55�� 59� -2(&�%.&) +� /���������� 3� *������ /� $��� R� /����� ����� ���� .��� �550� 2$

:$-;� -8&&�%.-) R� �������� R��+� 3��������� 5�� 0��� ���+� �� �������� ���8� �� 0� D�

$�� � �� 0����� @���������� +�� �541�%.I) +� <������� +� <��������� <� @���� 5� ������ %� ����� ���� %���

�554� $5 :5;� I66�%..) +� <������� +� <��������� <� @���� ����� 5��� 0��� .)��� �!��� ��

0��� ���������� R� ��� &2?-- �� �� �555�

UUUUUUUUUUUUUUUUUUUUUU

����� ���� �������� ����� ��� @� &- �� >??��������K�������� 7 -881 4�*"9�:/0 :����� $��0 ; /� <$� � 4������ &11I

��� �����

Volume 1, Issue 2, 2007 Can We Be More Efficient in Oil and Gas Exploitation? A Review of the Shortcomings of Recovery Factor and the Need for an Open Worldwide Production Database.

Gioia Falcone, Assistant Professor, Texas A&M University, [email protected] Bob Harrison, Director, Soluzioni Idrocarburi s.r.l., [email protected] Catalin Teodoriu, Assistant Professor, Texas A&M University, [email protected]

Abstract

This paper critically reviews the concept of recovery factor (RF) of oil and gas fields. Although this simple parameter is used throughout the oil and gas industry, it is subject to misunderstanding and misuse. Besides changing continually through the producing life of a field, the estimate of RF is affected by geological uncertainty, inappropriate reserves reporting, technological shortcomings, commercial practices and political decisions. At present, the information necessary to fully evaluate RF is not unequivocally determined, audited or reported, which makes it impossible to produce consistent global field statistics. Based on the authors’ experience, the paper outlines the shortcomings of RF and suggests how they may be overcome. To promote clarity and transparency in RF calculations, a template for an open worldwide production database is proposed. Introduction

Accepted wisdom suggests that the higher an oil or gas fields’ value of RF, the more efficient the hydrocarbons have been produced from the reservoir. Optimising the recovery from a hydrocarbon field should be the common goal of both Governments and Operators, although increasing production levels at the right economic and political moment may prove too tempting to some. Hence, the use of RF as a yardstick to measure the performance of a reservoir (or the management of that reservoir) has some serious shortcomings. In order to use RF appropriately, it is important to understand what it is, how it is calculated and the uncertainty inherent to the parameters from which it is derived. The value of RF is defined as the ratio of the recoverable volume to the hydrocarbons originally in place (HOIP) over the course of a field’s economic life. Yet this seemingly trivial calculation has inherent uncertainty and can vary due to many reasons. Among the many factors that impact on the ultimate recovery from a field are: the geology of the reservoir; the properties of the reservoir fluids; the drive mechanism; the technology used to drill, complete and produce the field; the oil and gas prices. The RF of “tight” hydrocarbons reservoirs can be as low as 1%, but it can be as high as 80% for reservoirs of excellent porosity and permeability. In addition, for a given geology, enhanced oil recovery (EOR) methods can recover more oil from the same reservoir. It is common practice to differentiate between primary, secondary, and tertiary (or enhanced) oil recovery. With primary recovery, the natural potential of the reservoir drives the oil to the surface and this can be combined with well-bore artificial lift techniques, but only a small amount (~10%) of the HOIP can be recovered by this method. Injecting water or gas to displace oil is referred to as secondary recovery and this usually allows a recovery of 20-40%. The field characteristics that lead to high primary and secondary recovery are summarised in Table 1. Tertiary recovery may lead to recovering ~60% or more of HOIP, using methods such as:

• Thermal recovery, by adding heat to the reservoir fluids to make them more mobile.

• Gas injection for miscible sweep of the oil, is achieved by injecting flue gases or CO2 , which dissolve in the oil, reducing its viscosity and increasing its mobility.

• Chemical injection, either using polymers to “thicken” the injected water to increase its viscosity and so improve water-flood efficiency, or using surfactants, to improve the mobility of the oil droplets by reducing the surface tension.

Table 1: Main features of high-recovery fields. (Schulte, 2005).

The RF values quoted above are suggested by the U.S. Department of Energy and reflect the USA experience. The numbers do change for different geographical areas, e.g. the average RF in the UK North Sea for a water-flood oil field is around 45% and may be up to 60% in some fields. However, care must be taken before applying average RF values to all fields as heavy oil, tight gas, HP/HT fractured and deepwater reservoirs present special challenges and, therefore, different levels of recovery should be expected. The estimate of RF has an element of time dependency. When considering hydrocarbon reserves, it is important to distinguish between the reserves of fields that have already been abandoned and the reserves of fields that are either about to come on stream or in the early years of production. The evolution of uncertainty for reserves estimation for a generic field is shown in figure 1, where the uncertainty reduces as more information is gathered from the field, from the exploration and appraisal stage to abandonment. Figure 1: Uncertainty of reserves estimation decreases as field life progresses

Development

Abandonment

Range of RF

Estimates

High estimateHigh estimate

Low estimateLow estimate

Time

Production

Appraisal

Exploration

Hence, reserves do change over an asset’s life as does HOIP as more reservoir data is gathered during development drilling and production history is matched. Some analysts claim that RF does not change over time, but the authors disagree with this statement. Not only do the reserves change over the life of a field, the method used to compute reserves also changes. Figure 2 illustrates that different reserves determination methods (use of results from analogue fields, volumetric calculations, decline curve analysis, material balance and numerical simulation) are used at different stages of a field’s life. Figure 2: Reserves determination methods change through field life

As mentioned earlier, besides time dependency, there are many other factors affect the RF, which reflect geological risk, regulatory guidelines, technological shortcomings, commercial practices or political stances. Some of the more important factors are:

� Uncertainty in value of HOIP. � Definition of reserves and reporting standards. � Metering error(s) when measuring produced volumes. � Application of new technology to enhance well productivity. � Change in operatorship of the asset. � Change in business model used by operator. � Paucity of verified field and well data in the public domain.

The above factors, their effect on RF estimates and how their impact may be lessened are discussed in detail below. Factors Affecting Estimation of RF

The following factors influence RF; some implicitly, some explicitly, but they can all have a major impact on its estimate. Uncertainty in HOIP.

The recoverable volume will only be known when all the reserves from that field have all been produced, but the HOIP will probably never be known for certain. This is because the cumulative produced volume can be metered, but the HOIP must be estimated from seismic surveys, well

Exploration

Development

Study

MethodAnalogue Volumetric

Performance

Simulation studies

Material balance studies

Decline trend analysis

Appraisal

Decline

Abandonment

Production

logs, geological models and hydrocarbon samples. Thus, even after all reserves have been produced, the actual RF will have inherent uncertainty as the HOIP is estimated, not measured. Hence, the HOIP of a field can go up or down, depending on revisions of the original numbers and/or field extensions. In the case of field extensions, the extra reservoir volume may be of a better or worse quality than the original discovery, therefore the overall RF may be lower or higher than that originally predicted.

Definition of Reserves.

There is currently no unique way of defining and assessing reserves, although the Society of Petroleum Engineers (SPE) and the World Petroleum Congress (WPC) have jointly published a guideline (SPE-WPC, 2007) that purports to do just that. Neither is there a global standard for financial reporting purposes. The reserves may be reported in different ways, more conservative or more likely, and the RF will reflect the choice of reserves reporting. Some of the reserves databases that are available in the public domain will be discussed later in the paper.

Reserves are usually estimated according to a probabilistic approach, where a differentiation is made between proven, probable and possible reserves. In some parts of the world, companies only have to report proved reserves (referred to as 1P or P90). In other countries, the probable reserves (2P or P50) are issued. Finally, some classification systems are extended to include possible reserves (3P or P10). As the value of an oil and gas company is a direct function of its forecast reserves base, the same company may show completely different results depending on whether 1P, 2P or 3P numbers are used. Production profiles that are based on 2P reserves are more optimistic than the 1P, yet more conservative than the 3P. In asset sales, the 3P or upside reserves of a field development may be taken into consideration. Different available databases report different types of reserves. This will be covered later in the paper. Figure 3 shows how reserve growth occurs when reserves are reported as proven, but does not statistically occur when the reserves are reported as probable (2P). The figure highlights the possible confusion that may arise in quantifying reserves when using 1P numbers only.

Figure 3: World remaining conventional oil & gas reserves from “political” & “technical” sources (Laherrère, 2006)

When aggregating the reserves of a company, two common practices exist. The first is based on the arithmetic summation of deterministic estimates, whilst the second performs a probabilistic (or statistical) aggregation of probabilistic distributions. The difference between these two methods is illustrated in Figure 4, which is taken from the guideline (SPE-WPC, 2007).

In order to correctly add together the ranges of reserves from a number of fields, the second method of probabilistic aggregation must be used. Reserves distributions tend to be skewed log-normal, as they are based on permeability variations. The only point where the deterministic and probabilistic results coincide is at the mean or average value of the aggregated distribution. Thus, when working out the overall reserves distribution curve of several fields, the P50 of the total curve does not correspond to the sum of the individual P50s. The log normal distribution ensures that the computed mean is always higher than the P50 value. The larger the skew of the resultant reserves distribution, the bigger the difference between the computed mean and the P50 value. Traditionally, companies have quoted reserves ranges as 1P-2P-3P (P90-P50-P10) values. However, the authors have seen many instances where the aggregated reserves of a number of fields have been incorrectly determined. It is the authors’ opinion that reporting a field or aggregated fields’ mean reserves over time are the best estimate for ultimate recovery. From the discussions above it follows that a unique definition of reserves is needed prior to defining the RF.

Figure 4: Deterministic versus Probabilistic Aggregation of Reserves (SPE-WPC, 2007).

Metering Error(s). When using reservoir modelling techniques to forecast oil and gas production, from which the ultimate field recovery can be predicted, the volumes and flow rates of fluids produced from a reservoir are used to tune the models. However, the metering of the produced fluids is not error free; the measurements may be taken with different levels of accuracy, depending on whether they are required for fiscal, allocation or reservoir management purposes. In the latter case, an accuracy of ±10% for the measurement of the produced hydrocarbons is generally considered to be acceptable. The metering uncertainty is particularly important for small discoveries or marginal fields, where the effect of wrongly predicting the ultimate reserves and RF can severely impact the overall field economics. Since the results from production measurements are implemented in the reservoir modelling or production optimisation processes, it is clear that the accuracy of such measurements will affect the prediction of ultimate recovery from a reservoir. More accurate measurements imply that this uncertainty can be reduced. It is also clear that different levels of uncertainty may be acceptable, depending on overall field reserves, oil price, production lifetime, etc.

Application of New Technology. It is often suggested that new technology can enhance the produced volumes and therefore accelerate and, in many cases, increase the recovery from a field. In the last decade, some fundamental technological advances have been made in remote detection (higher definition and

4D seismic, controlled source electro magnetic surveys (CSEM)), drilling and completions (e.g. geo-steering, multilateral producers, under-balanced drilling, smart wells), well logging, real time reservoir monitoring (including multiphase flow metering and fibre optics), sub-sea and down-hole technology (e.g. water shut-offs, water separation), multiphase transport (including multiphase pumps and wet gas compressors) and flow assurance. An example of where new technology has enhanced recoverable reserves is represented by tight gas reservoirs. These formations, which are classified as having permeability less than 0.1 md, may contain at least 1,300 x10

9 m

3 of gas worldwide (BGR, 1999). Many tight gas reservoirs were

discovered years ago, but their potential has still to be fully realised due to their low productivity when developed with “conventional” completions and reservoir management techniques. However, this situation is already changing, as new technology improvements and the recent gas price hikes have seen a rapid development of tight gas sands worldwide, and particularly in the US. Table 2 summarises the quantitative impact of new technologies on tight gas resources (Perry et al., 1998). The cumulative impact of new technology results in an average tight gas producer with an increased recovery of 25% and a cost reduction of 17%.

New Technology Impact Well Cost Impact Geo-steering for zone selection Keep well in higher quality pay zones + 5 %

Under Balanced Drilling (UBD) Less fluid loss, minimise formation damage + 5 %

Drill multiple wells from a single location Smaller footprint, quicker rig up/down - 5 %

All waste disposed of on-site Cuttings, drill fluids, produced water re-injected - 5 %

Better fracture materials management Bulk purchasing & handling, cheaper treatment - 25 %

High angle drilling Maximises reservoir interval penetrated + 5 %

Coiled Tubing drilling, CT also be used as production tubing - 10 %

Improved hydraulic fracture conductivity More complete clean-up + 40 %

Monthly operating costs - 20%

Table 2: Impact of new technology on tight gas recovery (after Perry et al, 1998)

With conventional technology, the RF for tight gas reservoirs is very low, up to 10%, which makes the majority of them uneconomic. However, new technology can increase tight gas RF to between 30 and 50% (Friedel, 2004). Combining several new technologies together has proved particularly successful in unlocking extra reserves from tight gas reservoirs such. One example of “combo-technology” is to geo-steer horizontal wellbores into better quality reservoir units, while avoiding depleted zones or aquifers, and subsequently completing them with multiple fracturing techniques (Bencic, 2005). The combined benefits are illustrated in Figure 5, which shows an increase in recovery of an additional 1000 million cubic metres of gas (at normal conditions) from a tight Permian sandstone.

Figure 5. Unlocking gas well reserves by using combined drilling and fracturing techniques. (Bencic, 2005) [GP=Gas Produced; P/z=Reservoir pressure/gas compressibility factor]

However, while it is true to say that advances in key technologies may allow more recovery from a field, it is also true that novel technology may sometimes only accelerate the production of oil and gas volumes that would be equally produced using older techniques. There also seems to be a correlation between RF values and oil and gas prices, and the higher the price, the more likely it is that new techniques are implemented in the field. There tends to be a lag time between when the oil and gas prices rise and when more expensive field development methods are adopted. An example of how the RF and the oil and gas prices are related is given by Canada’s tar sands. Canada was not a significant player in the oil reserves charts until the oil prices jumped to over 50 $/barrel. At that point, the investors felt more confident in trying new exploration and production techniques to increase the recovery factor of tar sands and, since then, the World Oil Statistics position Canada in second place after Saudi Arabia for reserves. Such dependencies of the RF could be easily proved (or not) if the appropriate data were made available by government agencies, as will be discussed later.

Change of Asset Ownership.

In order to quickly and efficiently implement field development opportunities, it is important to guarantee that the asset is always in good hands. Asset sales imply that a new study is carried out to evaluate the asset value, resulting in new investment and additional production. This is illustrated by an example in Figure 6, which shows the production profile for the Forties field in the UK. The observed increase in production resulted from the transfer of the asset from the previous operator to a new one, from BP to Apache in this case. Prior to its sale in 2003, the Forties field was stated to have HOIP of 4.2 billion barrels, but this was increased, after re-processing of 3D seismic and an aggressive infill drilling programme, to 5 billion barrels. The subsequent RF fell from the 2003 figure of 62% to 53%. Although the Forties field is a success case for change of asset ownership, such mature asset transfers can be financially complex due to decommissioning costs and liabilities (PILOT, 2005).

Figure 6: Forties field production, showing improvement (actual and forecast) over baseline by infill drilling programme by new operator (UK dti’s PPRS online database, 2006).

Change of Business Model. Another way to ensure the recovery from a field is maximised is to apply new business models and enhance cross-industry partnerships (PILOT, 2005). New start-up companies have appeared on the scene that invest in development projects and take technical control of them. They form

0

100

200

300

400

500

600

0 5 10 15 20 25 30 35 40 45 50

Years on stream

MBOPD

0

500

1,000

1,500

2,000

2,500

3,000

Cum M

MBO

Act Mb/d Baseline Forecast Rate Workover Profile

Act Cum MMb Baseline Forecast Prodn Workover Forecast Prodn

partnerships with field owners to help increase production and earn a share of the incremental revenue. A typical cash flow for such business models is shown schematically in Figure 7. Figure 7: Cash flow of production sharing business model. (EDP 2005)

Paucity of Accessible Production Databases

In order to investigate the values of the RF of hydrocarbon fields worldwide, it is necessary to review historical production data and forecast reserves from available databases. While some countries and states have their own collection of production data by field (or even better, by well), no “official” database exists for all producing fields and/or wells worldwide. In fact, many countries treat their figures as confidential and do not disclose them. Some of the resources that are available in the public domain include:

� BP Statistical Review of World Energy, summarise the production and consumption figures of hydrocarbon resources by country each year.

� Oil & Gas Journal (OGJ), provides production data down to an oil field level for each country in the world. This annual survey gives each oil field’s rate in barrels per day, the oil’s API gravity and the depth of the reservoir.

� Official online production databases, maintained and provided by some government and federal agencies. Some examples of these excellent, but limited resources are: o Well and field databases of some of the oil and gas producing states in the USA

(California, Wyoming, etc.) are exemplary, giving almost everything analysts need to compute RF with some confidence.

o Petroleum Production Reporting System (PPRS) of the UK’s Department of Trade and Industry (dti), which gives online monthly data for each oil and gas field in the UKCS only three months in arrears. [PPRS did provide detailed monthly production data on a well-by-well basis, but unfortunately this excellent service was discontinued in 1999 - a serious mistake in the authors’ view].

� Commercial Production & Reserves Databases, which attempt to provide a worldwide dataset, but they can only publish field data for the countries that disclose them. Two of the major providers of this data service are: o IHS, whose global database provides both reserves and HOIP figures, allowing

RF to be estimated (although its accuracy will depend on the accuracy of the input data, which is provided by the host governments and operating companies). IHS has a tendency to be more exploration focused in its reporting.

o Wood Mackenzie, whose global database only provides reserves and is less extensive than that of IHS, but covers some regions in greater depth. Wood Mackenzie has a tendency to be more production focused in its reporting.

Company Cash Flow

Partner Cash Flow

Project Opex

Because of these inconsistencies in the way production and reserves data are released and published, it is not unusual to discover different databases with different production and reserves figures for the same field. After discussing the various factors that impact on RF estimates, let us now see how their variability around the world makes it difficult to generate useful RF statistics without the availability of detailed and consistent field data from government agencies. RF Statistics and their Limitations

A note of caution is warranted for those who analyse and use production and reserves data in the public domain. Because of the inconsistencies in the way production and reserves data are released and published, it is very difficult to generate valid and consistent RF statistics across the world. There is also missing information from published databases, which is not being released by all countries and/or field owners worldwide. Nevertheless, several researchers have studied the published global data in an attempt to determine the range and average value of RF. Schulte (2005) reports a global average RF value of 34%; the results of his analysis are shown in Figure 8.

Figure 8: Global cumulative oil in place ordered for RF (data from IHS, no year specified). Schulte (2005).

Laherrère (2006) reports a slightly lower global average value of RF of 27%, as shown in Figure 9. It is unfortunate that this analysis does not provide the average field RF for 2001 and 2006 using the same fields, which would have been a better indicator of whether technology had improved RF over time.

Figure 9: Cumulative number of oil fields worldwide (less US onshore) ordered by their RF (data reported in IHS database for 2001 and 2006) (Laherrère, 2006).

RF (%)

Figures 8 and 9 indicate an average oil RF of around 30%, but it must be noted that such plots do not make any differentiation by oil quality (heavy vs. light), reservoir depth, location, water depth (if offshore), geology, drive mechanism, etc. Also, the information displayed does not touch on the technology used (if any) to improve the recovery. Another word of caution about statistics of the type shown in Figures 8 and 9: the stated RF could actually be higher if the oil originally in place has been overestimated and vice versa. In other words, it is impossible to separate the concept of recovery from that of resources originally in place, yet the latter are rarely published by government agencies. Template for an Open Access Worldwide Production Database

As shown above, there is an inherent inaccuracy in estimating, measuring and reporting all the information necessary to determine the RF of a field. Also, depending on whether the RF is being estimated at the pre-development stage of a field or during its producing life, different data are needed to evaluate the RF. For pre-development, analogue fields help with the estimate of the new discovery’s performance. In this case, the main criteria for screening analogue fields are the reservoir geology and fluid properties and the technology to be implemented. After field start-up, production data will become available and so allow the use of more sophisticated techniques such as decline curve analysis, material balance and/or numerical modelling to predict the field’s recovery. Accessibility to information on wells’ performance, field behaviour, HOIP, geology, fluid properties and technology adopted is essential when estimating the RF of a field. This is true whether the field is a new discovery or on production. The data required to determine the RF through a cascade of input and output variables are summarised in Table 3.

Item to be Reported Reported Variables for Item Outputs Derived from Variables

1a Monthly well producer records

gas-oil-water volumes, days online, WHP,

WHT, choke %

Producer Uptime, Well producing GOR-WOR-Water Cut, Well Decline Curve,

Reserves per well, Field Reserves

1b Monthly well injector records gas, water volumes, days online, WHP, WHTInjector Uptime, Injectivity model, Producer:Injector Ratio

1c Well Technology

Artificially lifted? Stimulated? Water shut-offs? Horizontal or vertical? Completion

details Aids selection of appropriate analogue fields

3a Monthly Field production records gas-oil-water volumesField producing GOR-WOR-Water Cut, Field Decline Curve, Field Reserves

3b Monthly Field injection records gas, water volumes Field Injectivity model (aquifer strength)

3c Field Technology

Gas compression? Multiphase pumping? EOR? Aids selection of appropriate analogue fields

4 Bottom hole Surveys

average reservoir pressure & temperature history (date and depth datum) Material balance for HOIP

5 Top & Base reservoir structure maps

depth contours, scale, well locations, fluid

contacts, major faults

Gross rock volume, Field area, Hydrocarbon

fill factor

6 Field Geological description

sandstone or carbonate, matrix porosity or naturally fractured, massive or thin-bedded Aids selection of appropriate analogue fields

7 Field Petrophysical parameters

porosity, water saturation, gross thickness,

net-to-gross

Hydrocarbon pore volume (at reservoir

conditions) [Combining Items 5 & 7]

8 Field PVT properties (Oil)

API gravity, solution GOR, viscosity, bubble point, FVF

STOIIP (at surface conditions) [Combining Items 5, 7 & 8]

9 Field PVT properties (Gas)

Gas gravity, condensate gravity, CGR, viscosity, dew point, H2S-CO2-N2, FVF

GIIP (at reservoir conditions) [Combining Items 5, 7 & 9]

10 Recovery Factor

Field Reserves (from 1a or 3a), HOIP

(from 8 or 9), with date reference Table 3: Template for what production data should be provided by government agencies. [WHP=Well Head Pressure; WHT=Well Head Temperature; GOR=Gas-Oil Ratio; WOR=Water-Oil Ratio; EOR=Enhanced Oil Recovery; PVT=Pressure-Volume-Temperature; FVF=Formation Volume Factor; CGR=Condensate-Gas Ratio; STOIIP=Stock Tank Oil Initially In Place]

Table 3 shows that field production data, injection data and reserves alone (which are the only data published in the majority of the available databases) are insufficient to estimate the RF from an engineering standpoint. Many more data are required, without which the published RF estimates must be treated with caution. Conclusions

• This paper has highlighted the uncertainty in RF determination for hydrocarbon fields. This

uncertainty is often underestimated and is due to the uncertainty in reserves determination and HOIP estimate.

• The concepts of RF and reserves are directly linked, so it is necessary to refer to the same reserves definition prior to being able to compare the values of RF worldwide in a consistent manner.

• HOIP should be published by government agencies, in order to avoid ambiguities when producing worldwide statistics.

• The publication of reliable HOIP and reserves will allow more confident estimates of RF to be made, which will help with the evaluation of new prospects by comparing them with analogue fields.

• Because of the strong interdependence between RF, reserves and HOIP estimates, it remains impossible to state, on a general basis, that new technology will improve the ultimate recovery from a field. This generalisation is particularly difficult considering that the geology of a reservoir, the properties of the fluids therein and its drive mechanism all impact on the estimate of RF.

• The authors believe that the stated dependencies of the RF could be easily proved (or not) if appropriate data were made available from government agencies. The data should include geology, fluid properties, well-by-well information, HOIP and a description of the type of drive mechanism and technology adopted for each field in the world. Although this may sound ambitious, it is based on the recognition that RF estimates issued without such supporting information may be totally meaningless.

References

Apache Corporation, (2005), Annual Report 2005.

Apache Corporation, (2005), presentation for Merrill Lynch, November 2005.

Bencic, A., (2005) “Hydraulic Fracturing of the Rotliegend Sandstone in N-Germany –Technology, Company History and Strategic Importance”, SPE Technology Transfer Workshop, SUCO 2005.

BGR, (1999) (Bundesanstalt für Geowissenschaften und Rohstoffe): „Reserven, Ressourcen und Verfügbarkeit von Energierohstoffen 1998“, Rohstoffwirtschaftliche Länderstudien, Band XVII: 400 S., Hannover, 1999.

dti, (2005) “Brown Book – UKCS Reserves”, published annually (2005)

EDP, (2005), presentation to PESGB Aberdeen, November 2005.

Friedel, T., (2004), “Numerical Simulation of Production from Tight Gas Reservoirs by Advanced Stimulation Technologies”, PhD thesis, TU Bergakademie Freiberg, 2004.

Laherrère, J., (1997), “Distribution and evolution of Recovery Factor”, Oil reserves conference, Paris, France, 11 November 1997.

Laherrère, J., (2006), “Oil and gas: what future?”, Groningen Annual Energy Convention, 21 November 2006.

Perry, K.F., Cleary, M.P., Curtiss, J.B., (1998), “New technology for tight gas sand”, World Energy Congress, Houston, USA, 13-18 September 1998.

PILOT, (2005), “Maximising Economic Recovery of the UK’s Oil and Gas Reserves”, report of PILOT 204, Brownfields Study, March 2005.

Schulte, W.M., (2005), “Challenges and Strategy for Increased Oil Recovery”, paper IPTC 10146, International Petroleum Technology Conference, Doha, Qatar, 21-23 November 2005.

SPE-WPC, (2007), “Petroleum Reserves and Resources Classification, Definitions, and Guidelines”, DRAFT for Industry Review, September 2006.


Recommended